Merge branch 'linus' into release

Conflicts:
	arch/x86/kernel/acpi/sleep.c

Signed-off-by: Len Brown <len.brown@intel.com>

17 files changed, 3278 insertions, 0 deletions

diff --git a/arch/unicore32/mm/Kconfig b/arch/unicore32/mm/Kconfig
new file mode 100644
index 00000000000..5f77fb3c63b
--- /dev/null
+++ b/arch/unicore32/mm/Kconfig
@@ -0,0 +1,50 @@
+comment "Processor Type"
+
+# Select CPU types depending on the architecture selected.  This selects
+# which CPUs we support in the kernel image, and the compiler instruction
+# optimiser behaviour.
+
+config CPU_UCV2
+	def_bool y
+
+comment "Processor Features"
+
+config CPU_ICACHE_DISABLE
+	bool "Disable I-Cache (I-bit)"
+	help
+	  Say Y here to disable the processor instruction cache. Unless
+	  you have a reason not to or are unsure, say N.
+
+config CPU_DCACHE_DISABLE
+	bool "Disable D-Cache (D-bit)"
+	help
+	  Say Y here to disable the processor data cache. Unless
+	  you have a reason not to or are unsure, say N.
+
+config CPU_DCACHE_WRITETHROUGH
+	bool "Force write through D-cache"
+	help
+	  Say Y here to use the data cache in writethrough mode. Unless you
+	  specifically require this or are unsure, say N.
+
+config CPU_DCACHE_LINE_DISABLE
+	bool "Disable D-cache line ops"
+	default y
+	help
+	  Say Y here to disable the data cache line operations.
+
+config CPU_TLB_SINGLE_ENTRY_DISABLE
+	bool "Disable TLB single entry ops"
+	default y
+	help
+	  Say Y here to disable the TLB single entry operations.
+
+config SWIOTLB
+	def_bool y
+
+config IOMMU_HELPER
+	def_bool SWIOTLB
+
+config NEED_SG_DMA_LENGTH
+	def_bool SWIOTLB
+
diff --git a/arch/unicore32/mm/Makefile b/arch/unicore32/mm/Makefile
new file mode 100644
index 00000000000..46c16669931
--- /dev/null
+++ b/arch/unicore32/mm/Makefile
@@ -0,0 +1,15 @@
+#
+# Makefile for the linux unicore-specific parts of the memory manager.
+#
+
+obj-y				:= extable.o fault.o init.o pgd.o mmu.o
+obj-y				+= flush.o ioremap.o
+
+obj-$(CONFIG_SWIOTLB)		+= dma-swiotlb.o
+
+obj-$(CONFIG_MODULES)		+= proc-syms.o
+
+obj-$(CONFIG_ALIGNMENT_TRAP)	+= alignment.o
+
+obj-$(CONFIG_CPU_UCV2)		+= cache-ucv2.o tlb-ucv2.o proc-ucv2.o
+
diff --git a/arch/unicore32/mm/alignment.c b/arch/unicore32/mm/alignment.c
new file mode 100644
index 00000000000..28f576d733e
--- /dev/null
+++ b/arch/unicore32/mm/alignment.c
@@ -0,0 +1,523 @@
+/*
+ * linux/arch/unicore32/mm/alignment.c
+ *
+ * Code specific to PKUnity SoC and UniCore ISA
+ *
+ * Copyright (C) 2001-2010 GUAN Xue-tao
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+/*
+ * TODO:
+ *  FPU ldm/stm not handling
+ */
+#include <linux/compiler.h>
+#include <linux/kernel.h>
+#include <linux/errno.h>
+#include <linux/string.h>
+#include <linux/init.h>
+#include <linux/sched.h>
+#include <linux/uaccess.h>
+
+#include <asm/tlbflush.h>
+#include <asm/unaligned.h>
+
+#define CODING_BITS(i)	(i & 0xe0000120)
+
+#define LDST_P_BIT(i)	(i & (1 << 28))	/* Preindex             */
+#define LDST_U_BIT(i)	(i & (1 << 27))	/* Add offset           */
+#define LDST_W_BIT(i)	(i & (1 << 25))	/* Writeback            */
+#define LDST_L_BIT(i)	(i & (1 << 24))	/* Load                 */
+
+#define LDST_P_EQ_U(i)	((((i) ^ ((i) >> 1)) & (1 << 27)) == 0)
+
+#define LDSTH_I_BIT(i)	(i & (1 << 26))	/* half-word immed      */
+#define LDM_S_BIT(i)	(i & (1 << 26))	/* write ASR from BSR */
+#define LDM_H_BIT(i)	(i & (1 << 6))	/* select r0-r15 or r16-r31 */
+
+#define RN_BITS(i)	((i >> 19) & 31)	/* Rn                   */
+#define RD_BITS(i)	((i >> 14) & 31)	/* Rd                   */
+#define RM_BITS(i)	(i & 31)	/* Rm                   */
+
+#define REGMASK_BITS(i)	(((i & 0x7fe00) >> 3) | (i & 0x3f))
+#define OFFSET_BITS(i)	(i & 0x03fff)
+
+#define SHIFT_BITS(i)	((i >> 9) & 0x1f)
+#define SHIFT_TYPE(i)	(i & 0xc0)
+#define SHIFT_LSL	0x00
+#define SHIFT_LSR	0x40
+#define SHIFT_ASR	0x80
+#define SHIFT_RORRRX	0xc0
+
+union offset_union {
+	unsigned long un;
+	signed long sn;
+};
+
+#define TYPE_ERROR	0
+#define TYPE_FAULT	1
+#define TYPE_LDST	2
+#define TYPE_DONE	3
+#define TYPE_SWAP  4
+#define TYPE_COLS  5		/* Coprocessor load/store */
+
+#define get8_unaligned_check(val, addr, err)		\
+	__asm__(					\
+	"1:	ldb.u	%1, [%2], #1\n"			\
+	"2:\n"						\
+	"	.pushsection .fixup,\"ax\"\n"		\
+	"	.align	2\n"				\
+	"3:	mov	%0, #1\n"			\
+	"	b	2b\n"				\
+	"	.popsection\n"				\
+	"	.pushsection __ex_table,\"a\"\n"		\
+	"	.align	3\n"				\
+	"	.long	1b, 3b\n"			\
+	"	.popsection\n"				\
+	: "=r" (err), "=&r" (val), "=r" (addr)		\
+	: "0" (err), "2" (addr))
+
+#define get8t_unaligned_check(val, addr, err)		\
+	__asm__(					\
+	"1:	ldb.u	%1, [%2], #1\n"			\
+	"2:\n"						\
+	"	.pushsection .fixup,\"ax\"\n"		\
+	"	.align	2\n"				\
+	"3:	mov	%0, #1\n"			\
+	"	b	2b\n"				\
+	"	.popsection\n"				\
+	"	.pushsection __ex_table,\"a\"\n"		\
+	"	.align	3\n"				\
+	"	.long	1b, 3b\n"			\
+	"	.popsection\n"				\
+	: "=r" (err), "=&r" (val), "=r" (addr)		\
+	: "0" (err), "2" (addr))
+
+#define get16_unaligned_check(val, addr)			\
+	do {							\
+		unsigned int err = 0, v, a = addr;		\
+		get8_unaligned_check(val, a, err);		\
+		get8_unaligned_check(v, a, err);		\
+		val |= v << 8;					\
+		if (err)					\
+			goto fault;				\
+	} while (0)
+
+#define put16_unaligned_check(val, addr)			\
+	do {							\
+		unsigned int err = 0, v = val, a = addr;	\
+		__asm__(					\
+		"1:	stb.u	%1, [%2], #1\n"			\
+		"	mov	%1, %1 >> #8\n"			\
+		"2:	stb.u	%1, [%2]\n"			\
+		"3:\n"						\
+		"	.pushsection .fixup,\"ax\"\n"		\
+		"	.align	2\n"				\
+		"4:	mov	%0, #1\n"			\
+		"	b	3b\n"				\
+		"	.popsection\n"				\
+		"	.pushsection __ex_table,\"a\"\n"		\
+		"	.align	3\n"				\
+		"	.long	1b, 4b\n"			\
+		"	.long	2b, 4b\n"			\
+		"	.popsection\n"				\
+		: "=r" (err), "=&r" (v), "=&r" (a)		\
+		: "0" (err), "1" (v), "2" (a));			\
+		if (err)					\
+			goto fault;				\
+	} while (0)
+
+#define __put32_unaligned_check(ins, val, addr)			\
+	do {							\
+		unsigned int err = 0, v = val, a = addr;	\
+		__asm__(					\
+		"1:	"ins"	%1, [%2], #1\n"			\
+		"	mov	%1, %1 >> #8\n"			\
+		"2:	"ins"	%1, [%2], #1\n"			\
+		"	mov	%1, %1 >> #8\n"			\
+		"3:	"ins"	%1, [%2], #1\n"			\
+		"	mov	%1, %1 >> #8\n"			\
+		"4:	"ins"	%1, [%2]\n"			\
+		"5:\n"						\
+		"	.pushsection .fixup,\"ax\"\n"		\
+		"	.align	2\n"				\
+		"6:	mov	%0, #1\n"			\
+		"	b	5b\n"				\
+		"	.popsection\n"				\
+		"	.pushsection __ex_table,\"a\"\n"		\
+		"	.align	3\n"				\
+		"	.long	1b, 6b\n"			\
+		"	.long	2b, 6b\n"			\
+		"	.long	3b, 6b\n"			\
+		"	.long	4b, 6b\n"			\
+		"	.popsection\n"				\
+		: "=r" (err), "=&r" (v), "=&r" (a)		\
+		: "0" (err), "1" (v), "2" (a));			\
+		if (err)					\
+			goto fault;				\
+	} while (0)
+
+#define get32_unaligned_check(val, addr)			\
+	do {							\
+		unsigned int err = 0, v, a = addr;		\
+		get8_unaligned_check(val, a, err);		\
+		get8_unaligned_check(v, a, err);		\
+		val |= v << 8;					\
+		get8_unaligned_check(v, a, err);		\
+		val |= v << 16;					\
+		get8_unaligned_check(v, a, err);		\
+		val |= v << 24;					\
+		if (err)					\
+			goto fault;				\
+	} while (0)
+
+#define put32_unaligned_check(val, addr)			\
+	__put32_unaligned_check("stb.u", val, addr)
+
+#define get32t_unaligned_check(val, addr)			\
+	do {							\
+		unsigned int err = 0, v, a = addr;		\
+		get8t_unaligned_check(val, a, err);		\
+		get8t_unaligned_check(v, a, err);		\
+		val |= v << 8;					\
+		get8t_unaligned_check(v, a, err);		\
+		val |= v << 16;					\
+		get8t_unaligned_check(v, a, err);		\
+		val |= v << 24;					\
+		if (err)					\
+			goto fault;				\
+	} while (0)
+
+#define put32t_unaligned_check(val, addr)			\
+	__put32_unaligned_check("stb.u", val, addr)
+
+static void
+do_alignment_finish_ldst(unsigned long addr, unsigned long instr,
+			 struct pt_regs *regs, union offset_union offset)
+{
+	if (!LDST_U_BIT(instr))
+		offset.un = -offset.un;
+
+	if (!LDST_P_BIT(instr))
+		addr += offset.un;
+
+	if (!LDST_P_BIT(instr) || LDST_W_BIT(instr))
+		regs->uregs[RN_BITS(instr)] = addr;
+}
+
+static int
+do_alignment_ldrhstrh(unsigned long addr, unsigned long instr,
+		      struct pt_regs *regs)
+{
+	unsigned int rd = RD_BITS(instr);
+
+	/* old value 0x40002120, can't judge swap instr correctly */
+	if ((instr & 0x4b003fe0) == 0x40000120)
+		goto swp;
+
+	if (LDST_L_BIT(instr)) {
+		unsigned long val;
+		get16_unaligned_check(val, addr);
+
+		/* signed half-word? */
+		if (instr & 0x80)
+			val = (signed long)((signed short)val);
+
+		regs->uregs[rd] = val;
+	} else
+		put16_unaligned_check(regs->uregs[rd], addr);
+
+	return TYPE_LDST;
+
+swp:
+	/* only handle swap word
+	 * for swap byte should not active this alignment exception */
+	get32_unaligned_check(regs->uregs[RD_BITS(instr)], addr);
+	put32_unaligned_check(regs->uregs[RM_BITS(instr)], addr);
+	return TYPE_SWAP;
+
+fault:
+	return TYPE_FAULT;
+}
+
+static int
+do_alignment_ldrstr(unsigned long addr, unsigned long instr,
+		    struct pt_regs *regs)
+{
+	unsigned int rd = RD_BITS(instr);
+
+	if (!LDST_P_BIT(instr) && LDST_W_BIT(instr))
+		goto trans;
+
+	if (LDST_L_BIT(instr))
+		get32_unaligned_check(regs->uregs[rd], addr);
+	else
+		put32_unaligned_check(regs->uregs[rd], addr);
+	return TYPE_LDST;
+
+trans:
+	if (LDST_L_BIT(instr))
+		get32t_unaligned_check(regs->uregs[rd], addr);
+	else
+		put32t_unaligned_check(regs->uregs[rd], addr);
+	return TYPE_LDST;
+
+fault:
+	return TYPE_FAULT;
+}
+
+/*
+ * LDM/STM alignment handler.
+ *
+ * There are 4 variants of this instruction:
+ *
+ * B = rn pointer before instruction, A = rn pointer after instruction
+ *              ------ increasing address ----->
+ *	        |    | r0 | r1 | ... | rx |    |
+ * PU = 01             B                    A
+ * PU = 11        B                    A
+ * PU = 00        A                    B
+ * PU = 10             A                    B
+ */
+static int
+do_alignment_ldmstm(unsigned long addr, unsigned long instr,
+		    struct pt_regs *regs)
+{
+	unsigned int rd, rn, pc_correction, reg_correction, nr_regs, regbits;
+	unsigned long eaddr, newaddr;
+
+	if (LDM_S_BIT(instr))
+		goto bad;
+
+	pc_correction = 4;	/* processor implementation defined */
+
+	/* count the number of registers in the mask to be transferred */
+	nr_regs = hweight16(REGMASK_BITS(instr)) * 4;
+
+	rn = RN_BITS(instr);
+	newaddr = eaddr = regs->uregs[rn];
+
+	if (!LDST_U_BIT(instr))
+		nr_regs = -nr_regs;
+	newaddr += nr_regs;
+	if (!LDST_U_BIT(instr))
+		eaddr = newaddr;
+
+	if (LDST_P_EQ_U(instr))	/* U = P */
+		eaddr += 4;
+
+	/*
+	 * This is a "hint" - we already have eaddr worked out by the
+	 * processor for us.
+	 */
+	if (addr != eaddr) {
+		printk(KERN_ERR "LDMSTM: PC = %08lx, instr = %08lx, "
+		       "addr = %08lx, eaddr = %08lx\n",
+		       instruction_pointer(regs), instr, addr, eaddr);
+		show_regs(regs);
+	}
+
+	if (LDM_H_BIT(instr))
+		reg_correction = 0x10;
+	else
+		reg_correction = 0x00;
+
+	for (regbits = REGMASK_BITS(instr), rd = 0; regbits;
+	     regbits >>= 1, rd += 1)
+		if (regbits & 1) {
+			if (LDST_L_BIT(instr))
+				get32_unaligned_check(regs->
+					uregs[rd + reg_correction], eaddr);
+			else
+				put32_unaligned_check(regs->
+					uregs[rd + reg_correction], eaddr);
+			eaddr += 4;
+		}
+
+	if (LDST_W_BIT(instr))
+		regs->uregs[rn] = newaddr;
+	return TYPE_DONE;
+
+fault:
+	regs->UCreg_pc -= pc_correction;
+	return TYPE_FAULT;
+
+bad:
+	printk(KERN_ERR "Alignment trap: not handling ldm with s-bit set\n");
+	return TYPE_ERROR;
+}
+
+static int
+do_alignment(unsigned long addr, unsigned int error_code, struct pt_regs *regs)
+{
+	union offset_union offset;
+	unsigned long instr, instrptr;
+	int (*handler) (unsigned long addr, unsigned long instr,
+			struct pt_regs *regs);
+	unsigned int type;
+
+	instrptr = instruction_pointer(regs);
+	if (instrptr >= PAGE_OFFSET)
+		instr = *(unsigned long *)instrptr;
+	else {
+		__asm__ __volatile__(
+				"ldw.u	%0, [%1]\n"
+				: "=&r"(instr)
+				: "r"(instrptr));
+	}
+
+	regs->UCreg_pc += 4;
+
+	switch (CODING_BITS(instr)) {
+	case 0x40000120:	/* ldrh or strh */
+		if (LDSTH_I_BIT(instr))
+			offset.un = (instr & 0x3e00) >> 4 | (instr & 31);
+		else
+			offset.un = regs->uregs[RM_BITS(instr)];
+		handler = do_alignment_ldrhstrh;
+		break;
+
+	case 0x60000000:	/* ldr or str immediate */
+	case 0x60000100:	/* ldr or str immediate */
+	case 0x60000020:	/* ldr or str immediate */
+	case 0x60000120:	/* ldr or str immediate */
+		offset.un = OFFSET_BITS(instr);
+		handler = do_alignment_ldrstr;
+		break;
+
+	case 0x40000000:	/* ldr or str register */
+		offset.un = regs->uregs[RM_BITS(instr)];
+		{
+			unsigned int shiftval = SHIFT_BITS(instr);
+
+			switch (SHIFT_TYPE(instr)) {
+			case SHIFT_LSL:
+				offset.un <<= shiftval;
+				break;
+
+			case SHIFT_LSR:
+				offset.un >>= shiftval;
+				break;
+
+			case SHIFT_ASR:
+				offset.sn >>= shiftval;
+				break;
+
+			case SHIFT_RORRRX:
+				if (shiftval == 0) {
+					offset.un >>= 1;
+					if (regs->UCreg_asr & PSR_C_BIT)
+						offset.un |= 1 << 31;
+				} else
+					offset.un = offset.un >> shiftval |
+					    offset.un << (32 - shiftval);
+				break;
+			}
+		}
+		handler = do_alignment_ldrstr;
+		break;
+
+	case 0x80000000:	/* ldm or stm */
+	case 0x80000020:	/* ldm or stm */
+		handler = do_alignment_ldmstm;
+		break;
+
+	default:
+		goto bad;
+	}
+
+	type = handler(addr, instr, regs);
+
+	if (type == TYPE_ERROR || type == TYPE_FAULT)
+		goto bad_or_fault;
+
+	if (type == TYPE_LDST)
+		do_alignment_finish_ldst(addr, instr, regs, offset);
+
+	return 0;
+
+bad_or_fault:
+	if (type == TYPE_ERROR)
+		goto bad;
+	regs->UCreg_pc -= 4;
+	/*
+	 * We got a fault - fix it up, or die.
+	 */
+	do_bad_area(addr, error_code, regs);
+	return 0;
+
+bad:
+	/*
+	 * Oops, we didn't handle the instruction.
+	 * However, we must handle fpu instr firstly.
+	 */
+#ifdef CONFIG_UNICORE_FPU_F64
+	/* handle co.load/store */
+#define CODING_COLS                0xc0000000
+#define COLS_OFFSET_BITS(i)	(i & 0x1FF)
+#define COLS_L_BITS(i)		(i & (1<<24))
+#define COLS_FN_BITS(i)		((i>>14) & 31)
+	if ((instr & 0xe0000000) == CODING_COLS) {
+		unsigned int fn = COLS_FN_BITS(instr);
+		unsigned long val = 0;
+		if (COLS_L_BITS(instr)) {
+			get32t_unaligned_check(val, addr);
+			switch (fn) {
+#define ASM_MTF(n)	case n:						\
+			__asm__ __volatile__("MTF %0, F" __stringify(n)	\
+				: : "r"(val));				\
+			break;
+			ASM_MTF(0); ASM_MTF(1); ASM_MTF(2); ASM_MTF(3);
+			ASM_MTF(4); ASM_MTF(5); ASM_MTF(6); ASM_MTF(7);
+			ASM_MTF(8); ASM_MTF(9); ASM_MTF(10); ASM_MTF(11);
+			ASM_MTF(12); ASM_MTF(13); ASM_MTF(14); ASM_MTF(15);
+			ASM_MTF(16); ASM_MTF(17); ASM_MTF(18); ASM_MTF(19);
+			ASM_MTF(20); ASM_MTF(21); ASM_MTF(22); ASM_MTF(23);
+			ASM_MTF(24); ASM_MTF(25); ASM_MTF(26); ASM_MTF(27);
+			ASM_MTF(28); ASM_MTF(29); ASM_MTF(30); ASM_MTF(31);
+#undef ASM_MTF
+			}
+		} else {
+			switch (fn) {
+#define ASM_MFF(n)	case n:						\
+			__asm__ __volatile__("MFF %0, F" __stringify(n)	\
+				: : "r"(val));				\
+			break;
+			ASM_MFF(0); ASM_MFF(1); ASM_MFF(2); ASM_MFF(3);
+			ASM_MFF(4); ASM_MFF(5); ASM_MFF(6); ASM_MFF(7);
+			ASM_MFF(8); ASM_MFF(9); ASM_MFF(10); ASM_MFF(11);
+			ASM_MFF(12); ASM_MFF(13); ASM_MFF(14); ASM_MFF(15);
+			ASM_MFF(16); ASM_MFF(17); ASM_MFF(18); ASM_MFF(19);
+			ASM_MFF(20); ASM_MFF(21); ASM_MFF(22); ASM_MFF(23);
+			ASM_MFF(24); ASM_MFF(25); ASM_MFF(26); ASM_MFF(27);
+			ASM_MFF(28); ASM_MFF(29); ASM_MFF(30); ASM_MFF(31);
+#undef ASM_MFF
+			}
+			put32t_unaligned_check(val, addr);
+		}
+		return TYPE_COLS;
+	}
+fault:
+	return TYPE_FAULT;
+#endif
+	printk(KERN_ERR "Alignment trap: not handling instruction "
+	       "%08lx at [<%08lx>]\n", instr, instrptr);
+	return 1;
+}
+
+/*
+ * This needs to be done after sysctl_init, otherwise sys/ will be
+ * overwritten.  Actually, this shouldn't be in sys/ at all since
+ * it isn't a sysctl, and it doesn't contain sysctl information.
+ */
+static int __init alignment_init(void)
+{
+	hook_fault_code(1, do_alignment, SIGBUS, BUS_ADRALN,
+			"alignment exception");
+
+	return 0;
+}
+
+fs_initcall(alignment_init);
diff --git a/arch/unicore32/mm/cache-ucv2.S b/arch/unicore32/mm/cache-ucv2.S
new file mode 100644
index 00000000000..ecaa1727f90
--- /dev/null
+++ b/arch/unicore32/mm/cache-ucv2.S
@@ -0,0 +1,212 @@
+/*
+ * linux/arch/unicore32/mm/cache-ucv2.S
+ *
+ * Code specific to PKUnity SoC and UniCore ISA
+ *
+ * Copyright (C) 2001-2010 GUAN Xue-tao
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ *  This is the "shell" of the UniCore-v2 processor support.
+ */
+#include <linux/linkage.h>
+#include <linux/init.h>
+#include <asm/assembler.h>
+#include <asm/page.h>
+
+#include "proc-macros.S"
+
+/*
+ *	__cpuc_flush_icache_all()
+ *	__cpuc_flush_kern_all()
+ *	__cpuc_flush_user_all()
+ *
+ *	Flush the entire cache.
+ */
+ENTRY(__cpuc_flush_icache_all)
+	/*FALLTHROUGH*/
+ENTRY(__cpuc_flush_kern_all)
+	/*FALLTHROUGH*/
+ENTRY(__cpuc_flush_user_all)
+	mov	r0, #0
+	movc	p0.c5, r0, #14			@ Dcache flush all
+	nop8
+
+	mov	r0, #0
+	movc	p0.c5, r0, #20			@ Icache invalidate all
+	nop8
+
+	mov	pc, lr
+
+/*
+ *	__cpuc_flush_user_range(start, end, flags)
+ *
+ *	Flush a range of TLB entries in the specified address space.
+ *
+ *	- start - start address (may not be aligned)
+ *	- end   - end address (exclusive, may not be aligned)
+ *	- flags	- vm_area_struct flags describing address space
+ */
+ENTRY(__cpuc_flush_user_range)
+	cxor.a	r2, #0
+	beq	__cpuc_dma_flush_range
+
+#ifndef CONFIG_CPU_DCACHE_LINE_DISABLE
+	andn	r0, r0, #CACHE_LINESIZE - 1	@ Safety check
+	sub	r1, r1, r0
+	csub.a	r1, #MAX_AREA_SIZE
+	bsg	2f
+
+	andn	r1, r1, #CACHE_LINESIZE - 1
+	add	r1, r1, #CACHE_LINESIZE
+
+101:	dcacheline_flush	r0, r11, r12
+
+	add	r0, r0, #CACHE_LINESIZE
+	sub.a	r1, r1, #CACHE_LINESIZE
+	bns	101b
+	b	3f
+#endif
+2:	mov	ip, #0
+	movc	p0.c5, ip, #14			@ Dcache flush all
+	nop8
+
+3:	mov	ip, #0
+	movc	p0.c5, ip, #20			@ Icache invalidate all
+	nop8
+
+	mov	pc, lr
+
+/*
+ *	__cpuc_coherent_kern_range(start,end)
+ *	__cpuc_coherent_user_range(start,end)
+ *
+ *	Ensure that the I and D caches are coherent within specified
+ *	region.  This is typically used when code has been written to
+ *	a memory region, and will be executed.
+ *
+ *	- start   - virtual start address of region
+ *	- end     - virtual end address of region
+ */
+ENTRY(__cpuc_coherent_kern_range)
+	/* FALLTHROUGH */
+ENTRY(__cpuc_coherent_user_range)
+#ifndef CONFIG_CPU_DCACHE_LINE_DISABLE
+	andn	r0, r0, #CACHE_LINESIZE - 1	@ Safety check
+	sub	r1, r1, r0
+	csub.a	r1, #MAX_AREA_SIZE
+	bsg	2f
+
+	andn	r1, r1, #CACHE_LINESIZE - 1
+	add	r1, r1, #CACHE_LINESIZE
+
+	@ r0 va2pa r10
+	mov	r9, #PAGE_SZ
+	sub	r9, r9, #1			@ PAGE_MASK
+101:	va2pa	r0, r10, r11, r12, r13, 2f	@ r10 is PA
+	b	103f
+102:	cand.a	r0, r9
+	beq	101b
+
+103:	movc	p0.c5, r10, #11			@ Dcache clean line of R10
+	nop8
+
+	add	r0, r0, #CACHE_LINESIZE
+	add	r10, r10, #CACHE_LINESIZE
+	sub.a	r1, r1, #CACHE_LINESIZE
+	bns	102b
+	b	3f
+#endif
+2:	mov	ip, #0
+	movc	p0.c5, ip, #10			@ Dcache clean all
+	nop8
+
+3:	mov	ip, #0
+	movc	p0.c5, ip, #20			@ Icache invalidate all
+	nop8
+
+	mov	pc, lr
+
+/*
+ *	__cpuc_flush_kern_dcache_area(void *addr, size_t size)
+ *
+ *	- addr	- kernel address
+ *	- size	- region size
+ */
+ENTRY(__cpuc_flush_kern_dcache_area)
+	mov	ip, #0
+	movc	p0.c5, ip, #14			@ Dcache flush all
+	nop8
+	mov	pc, lr
+
+/*
+ *	__cpuc_dma_clean_range(start,end)
+ *	- start   - virtual start address of region
+ *	- end     - virtual end address of region
+ */
+ENTRY(__cpuc_dma_clean_range)
+#ifndef CONFIG_CPU_DCACHE_LINE_DISABLE
+	andn	r0, r0, #CACHE_LINESIZE - 1
+	sub	r1, r1, r0
+	andn	r1, r1, #CACHE_LINESIZE - 1
+	add	r1, r1, #CACHE_LINESIZE
+
+	csub.a	r1, #MAX_AREA_SIZE
+	bsg	2f
+
+	@ r0 va2pa r10
+	mov	r9, #PAGE_SZ
+	sub	r9, r9, #1			@ PAGE_MASK
+101:	va2pa	r0, r10, r11, r12, r13, 2f	@ r10 is PA
+	b	1f
+102:	cand.a	r0, r9
+	beq	101b
+
+1:	movc	p0.c5, r10, #11			@ Dcache clean line of R10
+	nop8
+	add	r0, r0, #CACHE_LINESIZE
+	add	r10, r10, #CACHE_LINESIZE
+	sub.a	r1, r1, #CACHE_LINESIZE
+	bns	102b
+	mov	pc, lr
+#endif
+2:	mov	ip, #0
+	movc	p0.c5, ip, #10			@ Dcache clean all
+	nop8
+
+	mov	pc, lr
+
+/*
+ *	__cpuc_dma_inv_range(start,end)
+ *	__cpuc_dma_flush_range(start,end)
+ *	- start   - virtual start address of region
+ *	- end     - virtual end address of region
+ */
+__cpuc_dma_inv_range:
+	/* FALLTHROUGH */
+ENTRY(__cpuc_dma_flush_range)
+#ifndef CONFIG_CPU_DCACHE_LINE_DISABLE
+	andn	r0, r0, #CACHE_LINESIZE - 1
+	sub	r1, r1, r0
+	andn	r1, r1, #CACHE_LINESIZE - 1
+	add	r1, r1, #CACHE_LINESIZE
+
+	csub.a	r1, #MAX_AREA_SIZE
+	bsg	2f
+
+	@ r0 va2pa r10
+101:	dcacheline_flush	r0, r11, r12
+
+	add	r0, r0, #CACHE_LINESIZE
+	sub.a	r1, r1, #CACHE_LINESIZE
+	bns	101b
+	mov	pc, lr
+#endif
+2:	mov	ip, #0
+	movc	p0.c5, ip, #14			@ Dcache flush all
+	nop8
+
+	mov	pc, lr
+
diff --git a/arch/unicore32/mm/dma-swiotlb.c b/arch/unicore32/mm/dma-swiotlb.c
new file mode 100644
index 00000000000..bfa9fbb2bbb
--- /dev/null
+++ b/arch/unicore32/mm/dma-swiotlb.c
@@ -0,0 +1,34 @@
+/*
+ * Contains routines needed to support swiotlb for UniCore32.
+ *
+ * Copyright (C) 2010 Guan Xuetao
+ *
+ * This program is free software; you can redistribute  it and/or modify it
+ * under  the terms of  the GNU General  Public License as published by the
+ * Free Software Foundation;  either version 2 of the  License, or (at your
+ * option) any later version.
+ */
+#include <linux/pci.h>
+#include <linux/cache.h>
+#include <linux/module.h>
+#include <linux/dma-mapping.h>
+#include <linux/swiotlb.h>
+#include <linux/bootmem.h>
+
+#include <asm/dma.h>
+
+struct dma_map_ops swiotlb_dma_map_ops = {
+	.alloc_coherent = swiotlb_alloc_coherent,
+	.free_coherent = swiotlb_free_coherent,
+	.map_sg = swiotlb_map_sg_attrs,
+	.unmap_sg = swiotlb_unmap_sg_attrs,
+	.dma_supported = swiotlb_dma_supported,
+	.map_page = swiotlb_map_page,
+	.unmap_page = swiotlb_unmap_page,
+	.sync_single_for_cpu = swiotlb_sync_single_for_cpu,
+	.sync_single_for_device = swiotlb_sync_single_for_device,
+	.sync_sg_for_cpu = swiotlb_sync_sg_for_cpu,
+	.sync_sg_for_device = swiotlb_sync_sg_for_device,
+	.mapping_error = swiotlb_dma_mapping_error,
+};
+EXPORT_SYMBOL(swiotlb_dma_map_ops);
diff --git a/arch/unicore32/mm/extable.c b/arch/unicore32/mm/extable.c
new file mode 100644
index 00000000000..6564180eb28
--- /dev/null
+++ b/arch/unicore32/mm/extable.c
@@ -0,0 +1,24 @@
+/*
+ * linux/arch/unicore32/mm/extable.c
+ *
+ * Code specific to PKUnity SoC and UniCore ISA
+ *
+ * Copyright (C) 2001-2010 GUAN Xue-tao
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+#include <linux/module.h>
+#include <linux/uaccess.h>
+
+int fixup_exception(struct pt_regs *regs)
+{
+	const struct exception_table_entry *fixup;
+
+	fixup = search_exception_tables(instruction_pointer(regs));
+	if (fixup)
+		regs->UCreg_pc = fixup->fixup;
+
+	return fixup != NULL;
+}
diff --git a/arch/unicore32/mm/fault.c b/arch/unicore32/mm/fault.c
new file mode 100644
index 00000000000..283aa4b50b7
--- /dev/null
+++ b/arch/unicore32/mm/fault.c
@@ -0,0 +1,479 @@
+/*
+ * linux/arch/unicore32/mm/fault.c
+ *
+ * Code specific to PKUnity SoC and UniCore ISA
+ *
+ * Copyright (C) 2001-2010 GUAN Xue-tao
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+#include <linux/module.h>
+#include <linux/signal.h>
+#include <linux/mm.h>
+#include <linux/hardirq.h>
+#include <linux/init.h>
+#include <linux/kprobes.h>
+#include <linux/uaccess.h>
+#include <linux/page-flags.h>
+#include <linux/sched.h>
+#include <linux/io.h>
+
+#include <asm/system.h>
+#include <asm/pgtable.h>
+#include <asm/tlbflush.h>
+
+/*
+ * Fault status register encodings.  We steal bit 31 for our own purposes.
+ */
+#define FSR_LNX_PF		(1 << 31)
+
+static inline int fsr_fs(unsigned int fsr)
+{
+	/* xyabcde will be abcde+xy */
+	return (fsr & 31) + ((fsr & (3 << 5)) >> 5);
+}
+
+/*
+ * This is useful to dump out the page tables associated with
+ * 'addr' in mm 'mm'.
+ */
+void show_pte(struct mm_struct *mm, unsigned long addr)
+{
+	pgd_t *pgd;
+
+	if (!mm)
+		mm = &init_mm;
+
+	printk(KERN_ALERT "pgd = %p\n", mm->pgd);
+	pgd = pgd_offset(mm, addr);
+	printk(KERN_ALERT "[%08lx] *pgd=%08lx", addr, pgd_val(*pgd));
+
+	do {
+		pmd_t *pmd;
+		pte_t *pte;
+
+		if (pgd_none(*pgd))
+			break;
+
+		if (pgd_bad(*pgd)) {
+			printk("(bad)");
+			break;
+		}
+
+		pmd = pmd_offset((pud_t *) pgd, addr);
+		if (PTRS_PER_PMD != 1)
+			printk(", *pmd=%08lx", pmd_val(*pmd));
+
+		if (pmd_none(*pmd))
+			break;
+
+		if (pmd_bad(*pmd)) {
+			printk("(bad)");
+			break;
+		}
+
+		/* We must not map this if we have highmem enabled */
+		if (PageHighMem(pfn_to_page(pmd_val(*pmd) >> PAGE_SHIFT)))
+			break;
+
+		pte = pte_offset_map(pmd, addr);
+		printk(", *pte=%08lx", pte_val(*pte));
+		pte_unmap(pte);
+	} while (0);
+
+	printk("\n");
+}
+
+/*
+ * Oops.  The kernel tried to access some page that wasn't present.
+ */
+static void __do_kernel_fault(struct mm_struct *mm, unsigned long addr,
+		unsigned int fsr, struct pt_regs *regs)
+{
+	/*
+	 * Are we prepared to handle this kernel fault?
+	 */
+	if (fixup_exception(regs))
+		return;
+
+	/*
+	 * No handler, we'll have to terminate things with extreme prejudice.
+	 */
+	bust_spinlocks(1);
+	printk(KERN_ALERT
+	       "Unable to handle kernel %s at virtual address %08lx\n",
+	       (addr < PAGE_SIZE) ? "NULL pointer dereference" :
+	       "paging request", addr);
+
+	show_pte(mm, addr);
+	die("Oops", regs, fsr);
+	bust_spinlocks(0);
+	do_exit(SIGKILL);
+}
+
+/*
+ * Something tried to access memory that isn't in our memory map..
+ * User mode accesses just cause a SIGSEGV
+ */
+static void __do_user_fault(struct task_struct *tsk, unsigned long addr,
+		unsigned int fsr, unsigned int sig, int code,
+		struct pt_regs *regs)
+{
+	struct siginfo si;
+
+	tsk->thread.address = addr;
+	tsk->thread.error_code = fsr;
+	tsk->thread.trap_no = 14;
+	si.si_signo = sig;
+	si.si_errno = 0;
+	si.si_code = code;
+	si.si_addr = (void __user *)addr;
+	force_sig_info(sig, &si, tsk);
+}
+
+void do_bad_area(unsigned long addr, unsigned int fsr, struct pt_regs *regs)
+{
+	struct task_struct *tsk = current;
+	struct mm_struct *mm = tsk->active_mm;
+
+	/*
+	 * If we are in kernel mode at this point, we
+	 * have no context to handle this fault with.
+	 */
+	if (user_mode(regs))
+		__do_user_fault(tsk, addr, fsr, SIGSEGV, SEGV_MAPERR, regs);
+	else
+		__do_kernel_fault(mm, addr, fsr, regs);
+}
+
+#define VM_FAULT_BADMAP		0x010000
+#define VM_FAULT_BADACCESS	0x020000
+
+/*
+ * Check that the permissions on the VMA allow for the fault which occurred.
+ * If we encountered a write fault, we must have write permission, otherwise
+ * we allow any permission.
+ */
+static inline bool access_error(unsigned int fsr, struct vm_area_struct *vma)
+{
+	unsigned int mask = VM_READ | VM_WRITE | VM_EXEC;
+
+	if (!(fsr ^ 0x12))	/* write? */
+		mask = VM_WRITE;
+	if (fsr & FSR_LNX_PF)
+		mask = VM_EXEC;
+
+	return vma->vm_flags & mask ? false : true;
+}
+
+static int __do_pf(struct mm_struct *mm, unsigned long addr, unsigned int fsr,
+		struct task_struct *tsk)
+{
+	struct vm_area_struct *vma;
+	int fault;
+
+	vma = find_vma(mm, addr);
+	fault = VM_FAULT_BADMAP;
+	if (unlikely(!vma))
+		goto out;
+	if (unlikely(vma->vm_start > addr))
+		goto check_stack;
+
+	/*
+	 * Ok, we have a good vm_area for this
+	 * memory access, so we can handle it.
+	 */
+good_area:
+	if (access_error(fsr, vma)) {
+		fault = VM_FAULT_BADACCESS;
+		goto out;
+	}
+
+	/*
+	 * If for any reason at all we couldn't handle the fault, make
+	 * sure we exit gracefully rather than endlessly redo the fault.
+	 */
+	fault = handle_mm_fault(mm, vma, addr & PAGE_MASK,
+			    (!(fsr ^ 0x12)) ? FAULT_FLAG_WRITE : 0);
+	if (unlikely(fault & VM_FAULT_ERROR))
+		return fault;
+	if (fault & VM_FAULT_MAJOR)
+		tsk->maj_flt++;
+	else
+		tsk->min_flt++;
+	return fault;
+
+check_stack:
+	if (vma->vm_flags & VM_GROWSDOWN && !expand_stack(vma, addr))
+		goto good_area;
+out:
+	return fault;
+}
+
+static int do_pf(unsigned long addr, unsigned int fsr, struct pt_regs *regs)
+{
+	struct task_struct *tsk;
+	struct mm_struct *mm;
+	int fault, sig, code;
+
+	tsk = current;
+	mm = tsk->mm;
+
+	/*
+	 * If we're in an interrupt or have no user
+	 * context, we must not take the fault..
+	 */
+	if (in_atomic() || !mm)
+		goto no_context;
+
+	/*
+	 * As per x86, we may deadlock here.  However, since the kernel only
+	 * validly references user space from well defined areas of the code,
+	 * we can bug out early if this is from code which shouldn't.
+	 */
+	if (!down_read_trylock(&mm->mmap_sem)) {
+		if (!user_mode(regs)
+		    && !search_exception_tables(regs->UCreg_pc))
+			goto no_context;
+		down_read(&mm->mmap_sem);
+	} else {
+		/*
+		 * The above down_read_trylock() might have succeeded in
+		 * which case, we'll have missed the might_sleep() from
+		 * down_read()
+		 */
+		might_sleep();
+#ifdef CONFIG_DEBUG_VM
+		if (!user_mode(regs) &&
+		    !search_exception_tables(regs->UCreg_pc))
+			goto no_context;
+#endif
+	}
+
+	fault = __do_pf(mm, addr, fsr, tsk);
+	up_read(&mm->mmap_sem);
+
+	/*
+	 * Handle the "normal" case first - VM_FAULT_MAJOR / VM_FAULT_MINOR
+	 */
+	if (likely(!(fault &
+	       (VM_FAULT_ERROR | VM_FAULT_BADMAP | VM_FAULT_BADACCESS))))
+		return 0;
+
+	if (fault & VM_FAULT_OOM) {
+		/*
+		 * We ran out of memory, call the OOM killer, and return to
+		 * userspace (which will retry the fault, or kill us if we
+		 * got oom-killed)
+		 */
+		pagefault_out_of_memory();
+		return 0;
+	}
+
+	/*
+	 * If we are in kernel mode at this point, we
+	 * have no context to handle this fault with.
+	 */
+	if (!user_mode(regs))
+		goto no_context;
+
+	if (fault & VM_FAULT_SIGBUS) {
+		/*
+		 * We had some memory, but were unable to
+		 * successfully fix up this page fault.
+		 */
+		sig = SIGBUS;
+		code = BUS_ADRERR;
+	} else {
+		/*
+		 * Something tried to access memory that
+		 * isn't in our memory map..
+		 */
+		sig = SIGSEGV;
+		code = fault == VM_FAULT_BADACCESS ? SEGV_ACCERR : SEGV_MAPERR;
+	}
+
+	__do_user_fault(tsk, addr, fsr, sig, code, regs);
+	return 0;
+
+no_context:
+	__do_kernel_fault(mm, addr, fsr, regs);
+	return 0;
+}
+
+/*
+ * First Level Translation Fault Handler
+ *
+ * We enter here because the first level page table doesn't contain
+ * a valid entry for the address.
+ *
+ * If the address is in kernel space (>= TASK_SIZE), then we are
+ * probably faulting in the vmalloc() area.
+ *
+ * If the init_task's first level page tables contains the relevant
+ * entry, we copy the it to this task.  If not, we send the process
+ * a signal, fixup the exception, or oops the kernel.
+ *
+ * NOTE! We MUST NOT take any locks for this case. We may be in an
+ * interrupt or a critical region, and should only copy the information
+ * from the master page table, nothing more.
+ */
+static int do_ifault(unsigned long addr, unsigned int fsr, struct pt_regs *regs)
+{
+	unsigned int index;
+	pgd_t *pgd, *pgd_k;
+	pmd_t *pmd, *pmd_k;
+
+	if (addr < TASK_SIZE)
+		return do_pf(addr, fsr, regs);
+
+	if (user_mode(regs))
+		goto bad_area;
+
+	index = pgd_index(addr);
+
+	pgd = cpu_get_pgd() + index;
+	pgd_k = init_mm.pgd + index;
+
+	if (pgd_none(*pgd_k))
+		goto bad_area;
+
+	pmd_k = pmd_offset((pud_t *) pgd_k, addr);
+	pmd = pmd_offset((pud_t *) pgd, addr);
+
+	if (pmd_none(*pmd_k))
+		goto bad_area;
+
+	set_pmd(pmd, *pmd_k);
+	flush_pmd_entry(pmd);
+	return 0;
+
+bad_area:
+	do_bad_area(addr, fsr, regs);
+	return 0;
+}
+
+/*
+ * This abort handler always returns "fault".
+ */
+static int do_bad(unsigned long addr, unsigned int fsr, struct pt_regs *regs)
+{
+	return 1;
+}
+
+static int do_good(unsigned long addr, unsigned int fsr, struct pt_regs *regs)
+{
+	unsigned int res1, res2;
+
+	printk("dabt exception but no error!\n");
+
+	__asm__ __volatile__(
+			"mff %0,f0\n"
+			"mff %1,f1\n"
+			: "=r"(res1), "=r"(res2)
+			:
+			: "memory");
+
+	printk(KERN_EMERG "r0 :%08x  r1 :%08x\n", res1, res2);
+	panic("shut up\n");
+	return 0;
+}
+
+static struct fsr_info {
+	int (*fn) (unsigned long addr, unsigned int fsr, struct pt_regs *regs);
+	int sig;
+	int code;
+	const char *name;
+} fsr_info[] = {
+	/*
+	 * The following are the standard Unicore-I and UniCore-II aborts.
+	 */
+	{ do_good,	SIGBUS,  0,		"no error"		},
+	{ do_bad,	SIGBUS,  BUS_ADRALN,	"alignment exception"	},
+	{ do_bad,	SIGBUS,  BUS_OBJERR,	"external exception"	},
+	{ do_bad,	SIGBUS,  0,		"burst operation"	},
+	{ do_bad,	SIGBUS,  0,		"unknown 00100"		},
+	{ do_ifault,	SIGSEGV, SEGV_MAPERR,	"2nd level pt non-exist"},
+	{ do_bad,	SIGBUS,  0,		"2nd lvl large pt non-exist" },
+	{ do_bad,	SIGBUS,  0,		"invalid pte"		},
+	{ do_pf,	SIGSEGV, SEGV_MAPERR,	"page miss"		},
+	{ do_bad,	SIGBUS,  0,		"middle page miss"	},
+	{ do_bad,	SIGBUS,	 0,		"large page miss"	},
+	{ do_pf,	SIGSEGV, SEGV_MAPERR,	"super page (section) miss" },
+	{ do_bad,	SIGBUS,  0,		"unknown 01100"		},
+	{ do_bad,	SIGBUS,  0,		"unknown 01101"		},
+	{ do_bad,	SIGBUS,  0,		"unknown 01110"		},
+	{ do_bad,	SIGBUS,  0,		"unknown 01111"		},
+	{ do_bad,	SIGBUS,  0,		"addr: up 3G or IO"	},
+	{ do_pf,	SIGSEGV, SEGV_ACCERR,	"read unreadable addr"	},
+	{ do_pf,	SIGSEGV, SEGV_ACCERR,	"write unwriteable addr"},
+	{ do_pf,	SIGSEGV, SEGV_ACCERR,	"exec unexecutable addr"},
+	{ do_bad,	SIGBUS,  0,		"unknown 10100"		},
+	{ do_bad,	SIGBUS,  0,		"unknown 10101"		},
+	{ do_bad,	SIGBUS,  0,		"unknown 10110"		},
+	{ do_bad,	SIGBUS,  0,		"unknown 10111"		},
+	{ do_bad,	SIGBUS,  0,		"unknown 11000"		},
+	{ do_bad,	SIGBUS,  0,		"unknown 11001"		},
+	{ do_bad,	SIGBUS,  0,		"unknown 11010"		},
+	{ do_bad,	SIGBUS,  0,		"unknown 11011"		},
+	{ do_bad,	SIGBUS,  0,		"unknown 11100"		},
+	{ do_bad,	SIGBUS,  0,		"unknown 11101"		},
+	{ do_bad,	SIGBUS,  0,		"unknown 11110"		},
+	{ do_bad,	SIGBUS,  0,		"unknown 11111"		}
+};
+
+void __init hook_fault_code(int nr,
+		int (*fn) (unsigned long, unsigned int, struct pt_regs *),
+		int sig, int code, const char *name)
+{
+	if (nr < 0 || nr >= ARRAY_SIZE(fsr_info))
+		BUG();
+
+	fsr_info[nr].fn   = fn;
+	fsr_info[nr].sig  = sig;
+	fsr_info[nr].code = code;
+	fsr_info[nr].name = name;
+}
+
+/*
+ * Dispatch a data abort to the relevant handler.
+ */
+asmlinkage void do_DataAbort(unsigned long addr, unsigned int fsr,
+			struct pt_regs *regs)
+{
+	const struct fsr_info *inf = fsr_info + fsr_fs(fsr);
+	struct siginfo info;
+
+	if (!inf->fn(addr, fsr & ~FSR_LNX_PF, regs))
+		return;
+
+	printk(KERN_ALERT "Unhandled fault: %s (0x%03x) at 0x%08lx\n",
+	       inf->name, fsr, addr);
+
+	info.si_signo = inf->sig;
+	info.si_errno = 0;
+	info.si_code = inf->code;
+	info.si_addr = (void __user *)addr;
+	uc32_notify_die("", regs, &info, fsr, 0);
+}
+
+asmlinkage void do_PrefetchAbort(unsigned long addr,
+			unsigned int ifsr, struct pt_regs *regs)
+{
+	const struct fsr_info *inf = fsr_info + fsr_fs(ifsr);
+	struct siginfo info;
+
+	if (!inf->fn(addr, ifsr | FSR_LNX_PF, regs))
+		return;
+
+	printk(KERN_ALERT "Unhandled prefetch abort: %s (0x%03x) at 0x%08lx\n",
+	       inf->name, ifsr, addr);
+
+	info.si_signo = inf->sig;
+	info.si_errno = 0;
+	info.si_code = inf->code;
+	info.si_addr = (void __user *)addr;
+	uc32_notify_die("", regs, &info, ifsr, 0);
+}
diff --git a/arch/unicore32/mm/flush.c b/arch/unicore32/mm/flush.c
new file mode 100644
index 00000000000..93478cc8b26
--- /dev/null
+++ b/arch/unicore32/mm/flush.c
@@ -0,0 +1,98 @@
+/*
+ * linux/arch/unicore32/mm/flush.c
+ *
+ * Code specific to PKUnity SoC and UniCore ISA
+ *
+ * Copyright (C) 2001-2010 GUAN Xue-tao
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+#include <linux/module.h>
+#include <linux/mm.h>
+#include <linux/pagemap.h>
+
+#include <asm/cacheflush.h>
+#include <asm/system.h>
+#include <asm/tlbflush.h>
+
+void flush_cache_mm(struct mm_struct *mm)
+{
+}
+
+void flush_cache_range(struct vm_area_struct *vma, unsigned long start,
+		unsigned long end)
+{
+	if (vma->vm_flags & VM_EXEC)
+		__flush_icache_all();
+}
+
+void flush_cache_page(struct vm_area_struct *vma, unsigned long user_addr,
+		unsigned long pfn)
+{
+}
+
+static void flush_ptrace_access(struct vm_area_struct *vma, struct page *page,
+			 unsigned long uaddr, void *kaddr, unsigned long len)
+{
+	/* VIPT non-aliasing D-cache */
+	if (vma->vm_flags & VM_EXEC) {
+		unsigned long addr = (unsigned long)kaddr;
+
+		__cpuc_coherent_kern_range(addr, addr + len);
+	}
+}
+
+/*
+ * Copy user data from/to a page which is mapped into a different
+ * processes address space.  Really, we want to allow our "user
+ * space" model to handle this.
+ *
+ * Note that this code needs to run on the current CPU.
+ */
+void copy_to_user_page(struct vm_area_struct *vma, struct page *page,
+		       unsigned long uaddr, void *dst, const void *src,
+		       unsigned long len)
+{
+	memcpy(dst, src, len);
+	flush_ptrace_access(vma, page, uaddr, dst, len);
+}
+
+void __flush_dcache_page(struct address_space *mapping, struct page *page)
+{
+	/*
+	 * Writeback any data associated with the kernel mapping of this
+	 * page.  This ensures that data in the physical page is mutually
+	 * coherent with the kernels mapping.
+	 */
+	__cpuc_flush_kern_dcache_area(page_address(page), PAGE_SIZE);
+}
+
+/*
+ * Ensure cache coherency between kernel mapping and userspace mapping
+ * of this page.
+ */
+void flush_dcache_page(struct page *page)
+{
+	struct address_space *mapping;
+
+	/*
+	 * The zero page is never written to, so never has any dirty
+	 * cache lines, and therefore never needs to be flushed.
+	 */
+	if (page == ZERO_PAGE(0))
+		return;
+
+	mapping = page_mapping(page);
+
+	if (mapping && !mapping_mapped(mapping))
+		clear_bit(PG_dcache_clean, &page->flags);
+	else {
+		__flush_dcache_page(mapping, page);
+		if (mapping)
+			__flush_icache_all();
+		set_bit(PG_dcache_clean, &page->flags);
+	}
+}
+EXPORT_SYMBOL(flush_dcache_page);
diff --git a/arch/unicore32/mm/init.c b/arch/unicore32/mm/init.c
new file mode 100644
index 00000000000..3dbe3709b69
--- /dev/null
+++ b/arch/unicore32/mm/init.c
@@ -0,0 +1,517 @@
+/*
+ *  linux/arch/unicore32/mm/init.c
+ *
+ *  Copyright (C) 2010 GUAN Xue-tao
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+#include <linux/kernel.h>
+#include <linux/errno.h>
+#include <linux/swap.h>
+#include <linux/init.h>
+#include <linux/bootmem.h>
+#include <linux/mman.h>
+#include <linux/nodemask.h>
+#include <linux/initrd.h>
+#include <linux/highmem.h>
+#include <linux/gfp.h>
+#include <linux/memblock.h>
+#include <linux/sort.h>
+#include <linux/dma-mapping.h>
+
+#include <asm/sections.h>
+#include <asm/setup.h>
+#include <asm/sizes.h>
+#include <asm/tlb.h>
+#include <mach/map.h>
+
+#include "mm.h"
+
+static unsigned long phys_initrd_start __initdata = 0x01000000;
+static unsigned long phys_initrd_size __initdata = SZ_8M;
+
+static int __init early_initrd(char *p)
+{
+	unsigned long start, size;
+	char *endp;
+
+	start = memparse(p, &endp);
+	if (*endp == ',') {
+		size = memparse(endp + 1, NULL);
+
+		phys_initrd_start = start;
+		phys_initrd_size = size;
+	}
+	return 0;
+}
+early_param("initrd", early_initrd);
+
+/*
+ * This keeps memory configuration data used by a couple memory
+ * initialization functions, as well as show_mem() for the skipping
+ * of holes in the memory map.  It is populated by uc32_add_memory().
+ */
+struct meminfo meminfo;
+
+void show_mem(void)
+{
+	int free = 0, total = 0, reserved = 0;
+	int shared = 0, cached = 0, slab = 0, i;
+	struct meminfo *mi = &meminfo;
+
+	printk(KERN_DEFAULT "Mem-info:\n");
+	show_free_areas();
+
+	for_each_bank(i, mi) {
+		struct membank *bank = &mi->bank[i];
+		unsigned int pfn1, pfn2;
+		struct page *page, *end;
+
+		pfn1 = bank_pfn_start(bank);
+		pfn2 = bank_pfn_end(bank);
+
+		page = pfn_to_page(pfn1);
+		end  = pfn_to_page(pfn2 - 1) + 1;
+
+		do {
+			total++;
+			if (PageReserved(page))
+				reserved++;
+			else if (PageSwapCache(page))
+				cached++;
+			else if (PageSlab(page))
+				slab++;
+			else if (!page_count(page))
+				free++;
+			else
+				shared += page_count(page) - 1;
+			page++;
+		} while (page < end);
+	}
+
+	printk(KERN_DEFAULT "%d pages of RAM\n", total);
+	printk(KERN_DEFAULT "%d free pages\n", free);
+	printk(KERN_DEFAULT "%d reserved pages\n", reserved);
+	printk(KERN_DEFAULT "%d slab pages\n", slab);
+	printk(KERN_DEFAULT "%d pages shared\n", shared);
+	printk(KERN_DEFAULT "%d pages swap cached\n", cached);
+}
+
+static void __init find_limits(unsigned long *min, unsigned long *max_low,
+	unsigned long *max_high)
+{
+	struct meminfo *mi = &meminfo;
+	int i;
+
+	*min = -1UL;
+	*max_low = *max_high = 0;
+
+	for_each_bank(i, mi) {
+		struct membank *bank = &mi->bank[i];
+		unsigned long start, end;
+
+		start = bank_pfn_start(bank);
+		end = bank_pfn_end(bank);
+
+		if (*min > start)
+			*min = start;
+		if (*max_high < end)
+			*max_high = end;
+		if (bank->highmem)
+			continue;
+		if (*max_low < end)
+			*max_low = end;
+	}
+}
+
+static void __init uc32_bootmem_init(unsigned long start_pfn,
+	unsigned long end_pfn)
+{
+	struct memblock_region *reg;
+	unsigned int boot_pages;
+	phys_addr_t bitmap;
+	pg_data_t *pgdat;
+
+	/*
+	 * Allocate the bootmem bitmap page.  This must be in a region
+	 * of memory which has already been mapped.
+	 */
+	boot_pages = bootmem_bootmap_pages(end_pfn - start_pfn);
+	bitmap = memblock_alloc_base(boot_pages << PAGE_SHIFT, L1_CACHE_BYTES,
+				__pfn_to_phys(end_pfn));
+
+	/*
+	 * Initialise the bootmem allocator, handing the
+	 * memory banks over to bootmem.
+	 */
+	node_set_online(0);
+	pgdat = NODE_DATA(0);
+	init_bootmem_node(pgdat, __phys_to_pfn(bitmap), start_pfn, end_pfn);
+
+	/* Free the lowmem regions from memblock into bootmem. */
+	for_each_memblock(memory, reg) {
+		unsigned long start = memblock_region_memory_base_pfn(reg);
+		unsigned long end = memblock_region_memory_end_pfn(reg);
+
+		if (end >= end_pfn)
+			end = end_pfn;
+		if (start >= end)
+			break;
+
+		free_bootmem(__pfn_to_phys(start), (end - start) << PAGE_SHIFT);
+	}
+
+	/* Reserve the lowmem memblock reserved regions in bootmem. */
+	for_each_memblock(reserved, reg) {
+		unsigned long start = memblock_region_reserved_base_pfn(reg);
+		unsigned long end = memblock_region_reserved_end_pfn(reg);
+
+		if (end >= end_pfn)
+			end = end_pfn;
+		if (start >= end)
+			break;
+
+		reserve_bootmem(__pfn_to_phys(start),
+			(end - start) << PAGE_SHIFT, BOOTMEM_DEFAULT);
+	}
+}
+
+static void __init uc32_bootmem_free(unsigned long min, unsigned long max_low,
+	unsigned long max_high)
+{
+	unsigned long zone_size[MAX_NR_ZONES], zhole_size[MAX_NR_ZONES];
+	struct memblock_region *reg;
+
+	/*
+	 * initialise the zones.
+	 */
+	memset(zone_size, 0, sizeof(zone_size));
+
+	/*
+	 * The memory size has already been determined.  If we need
+	 * to do anything fancy with the allocation of this memory
+	 * to the zones, now is the time to do it.
+	 */
+	zone_size[0] = max_low - min;
+
+	/*
+	 * Calculate the size of the holes.
+	 *  holes = node_size - sum(bank_sizes)
+	 */
+	memcpy(zhole_size, zone_size, sizeof(zhole_size));
+	for_each_memblock(memory, reg) {
+		unsigned long start = memblock_region_memory_base_pfn(reg);
+		unsigned long end = memblock_region_memory_end_pfn(reg);
+
+		if (start < max_low) {
+			unsigned long low_end = min(end, max_low);
+			zhole_size[0] -= low_end - start;
+		}
+	}
+
+	/*
+	 * Adjust the sizes according to any special requirements for
+	 * this machine type.
+	 */
+	arch_adjust_zones(zone_size, zhole_size);
+
+	free_area_init_node(0, zone_size, min, zhole_size);
+}
+
+int pfn_valid(unsigned long pfn)
+{
+	return memblock_is_memory(pfn << PAGE_SHIFT);
+}
+EXPORT_SYMBOL(pfn_valid);
+
+static void uc32_memory_present(void)
+{
+}
+
+static int __init meminfo_cmp(const void *_a, const void *_b)
+{
+	const struct membank *a = _a, *b = _b;
+	long cmp = bank_pfn_start(a) - bank_pfn_start(b);
+	return cmp < 0 ? -1 : cmp > 0 ? 1 : 0;
+}
+
+void __init uc32_memblock_init(struct meminfo *mi)
+{
+	int i;
+
+	sort(&meminfo.bank, meminfo.nr_banks, sizeof(meminfo.bank[0]),
+		meminfo_cmp, NULL);
+
+	memblock_init();
+	for (i = 0; i < mi->nr_banks; i++)
+		memblock_add(mi->bank[i].start, mi->bank[i].size);
+
+	/* Register the kernel text, kernel data and initrd with memblock. */
+	memblock_reserve(__pa(_text), _end - _text);
+
+#ifdef CONFIG_BLK_DEV_INITRD
+	if (phys_initrd_size) {
+		memblock_reserve(phys_initrd_start, phys_initrd_size);
+
+		/* Now convert initrd to virtual addresses */
+		initrd_start = __phys_to_virt(phys_initrd_start);
+		initrd_end = initrd_start + phys_initrd_size;
+	}
+#endif
+
+	uc32_mm_memblock_reserve();
+
+	memblock_analyze();
+	memblock_dump_all();
+}
+
+void __init bootmem_init(void)
+{
+	unsigned long min, max_low, max_high;
+
+	max_low = max_high = 0;
+
+	find_limits(&min, &max_low, &max_high);
+
+	uc32_bootmem_init(min, max_low);
+
+#ifdef CONFIG_SWIOTLB
+	swiotlb_init(1);
+#endif
+	/*
+	 * Sparsemem tries to allocate bootmem in memory_present(),
+	 * so must be done after the fixed reservations
+	 */
+	uc32_memory_present();
+
+	/*
+	 * sparse_init() needs the bootmem allocator up and running.
+	 */
+	sparse_init();
+
+	/*
+	 * Now free the memory - free_area_init_node needs
+	 * the sparse mem_map arrays initialized by sparse_init()
+	 * for memmap_init_zone(), otherwise all PFNs are invalid.
+	 */
+	uc32_bootmem_free(min, max_low, max_high);
+
+	high_memory = __va((max_low << PAGE_SHIFT) - 1) + 1;
+
+	/*
+	 * This doesn't seem to be used by the Linux memory manager any
+	 * more, but is used by ll_rw_block.  If we can get rid of it, we
+	 * also get rid of some of the stuff above as well.
+	 *
+	 * Note: max_low_pfn and max_pfn reflect the number of _pages_ in
+	 * the system, not the maximum PFN.
+	 */
+	max_low_pfn = max_low - PHYS_PFN_OFFSET;
+	max_pfn = max_high - PHYS_PFN_OFFSET;
+}
+
+static inline int free_area(unsigned long pfn, unsigned long end, char *s)
+{
+	unsigned int pages = 0, size = (end - pfn) << (PAGE_SHIFT - 10);
+
+	for (; pfn < end; pfn++) {
+		struct page *page = pfn_to_page(pfn);
+		ClearPageReserved(page);
+		init_page_count(page);
+		__free_page(page);
+		pages++;
+	}
+
+	if (size && s)
+		printk(KERN_INFO "Freeing %s memory: %dK\n", s, size);
+
+	return pages;
+}
+
+static inline void
+free_memmap(unsigned long start_pfn, unsigned long end_pfn)
+{
+	struct page *start_pg, *end_pg;
+	unsigned long pg, pgend;
+
+	/*
+	 * Convert start_pfn/end_pfn to a struct page pointer.
+	 */
+	start_pg = pfn_to_page(start_pfn - 1) + 1;
+	end_pg = pfn_to_page(end_pfn);
+
+	/*
+	 * Convert to physical addresses, and
+	 * round start upwards and end downwards.
+	 */
+	pg = PAGE_ALIGN(__pa(start_pg));
+	pgend = __pa(end_pg) & PAGE_MASK;
+
+	/*
+	 * If there are free pages between these,
+	 * free the section of the memmap array.
+	 */
+	if (pg < pgend)
+		free_bootmem(pg, pgend - pg);
+}
+
+/*
+ * The mem_map array can get very big.  Free the unused area of the memory map.
+ */
+static void __init free_unused_memmap(struct meminfo *mi)
+{
+	unsigned long bank_start, prev_bank_end = 0;
+	unsigned int i;
+
+	/*
+	 * This relies on each bank being in address order.
+	 * The banks are sorted previously in bootmem_init().
+	 */
+	for_each_bank(i, mi) {
+		struct membank *bank = &mi->bank[i];
+
+		bank_start = bank_pfn_start(bank);
+
+		/*
+		 * If we had a previous bank, and there is a space
+		 * between the current bank and the previous, free it.
+		 */
+		if (prev_bank_end && prev_bank_end < bank_start)
+			free_memmap(prev_bank_end, bank_start);
+
+		/*
+		 * Align up here since the VM subsystem insists that the
+		 * memmap entries are valid from the bank end aligned to
+		 * MAX_ORDER_NR_PAGES.
+		 */
+		prev_bank_end = ALIGN(bank_pfn_end(bank), MAX_ORDER_NR_PAGES);
+	}
+}
+
+/*
+ * mem_init() marks the free areas in the mem_map and tells us how much
+ * memory is free.  This is done after various parts of the system have
+ * claimed their memory after the kernel image.
+ */
+void __init mem_init(void)
+{
+	unsigned long reserved_pages, free_pages;
+	struct memblock_region *reg;
+	int i;
+
+	max_mapnr   = pfn_to_page(max_pfn + PHYS_PFN_OFFSET) - mem_map;
+
+	/* this will put all unused low memory onto the freelists */
+	free_unused_memmap(&meminfo);
+
+	totalram_pages += free_all_bootmem();
+
+	reserved_pages = free_pages = 0;
+
+	for_each_bank(i, &meminfo) {
+		struct membank *bank = &meminfo.bank[i];
+		unsigned int pfn1, pfn2;
+		struct page *page, *end;
+
+		pfn1 = bank_pfn_start(bank);
+		pfn2 = bank_pfn_end(bank);
+
+		page = pfn_to_page(pfn1);
+		end  = pfn_to_page(pfn2 - 1) + 1;
+
+		do {
+			if (PageReserved(page))
+				reserved_pages++;
+			else if (!page_count(page))
+				free_pages++;
+			page++;
+		} while (page < end);
+	}
+
+	/*
+	 * Since our memory may not be contiguous, calculate the
+	 * real number of pages we have in this system
+	 */
+	printk(KERN_INFO "Memory:");
+	num_physpages = 0;
+	for_each_memblock(memory, reg) {
+		unsigned long pages = memblock_region_memory_end_pfn(reg) -
+			memblock_region_memory_base_pfn(reg);
+		num_physpages += pages;
+		printk(" %ldMB", pages >> (20 - PAGE_SHIFT));
+	}
+	printk(" = %luMB total\n", num_physpages >> (20 - PAGE_SHIFT));
+
+	printk(KERN_NOTICE "Memory: %luk/%luk available, %luk reserved, %luK highmem\n",
+		nr_free_pages() << (PAGE_SHIFT-10),
+		free_pages << (PAGE_SHIFT-10),
+		reserved_pages << (PAGE_SHIFT-10),
+		totalhigh_pages << (PAGE_SHIFT-10));
+
+	printk(KERN_NOTICE "Virtual kernel memory layout:\n"
+		"    vector  : 0x%08lx - 0x%08lx   (%4ld kB)\n"
+		"    vmalloc : 0x%08lx - 0x%08lx   (%4ld MB)\n"
+		"    lowmem  : 0x%08lx - 0x%08lx   (%4ld MB)\n"
+		"    modules : 0x%08lx - 0x%08lx   (%4ld MB)\n"
+		"      .init : 0x%p" " - 0x%p" "   (%4d kB)\n"
+		"      .text : 0x%p" " - 0x%p" "   (%4d kB)\n"
+		"      .data : 0x%p" " - 0x%p" "   (%4d kB)\n",
+
+		VECTORS_BASE, VECTORS_BASE + PAGE_SIZE,
+		DIV_ROUND_UP(PAGE_SIZE, SZ_1K),
+		VMALLOC_START, VMALLOC_END,
+		DIV_ROUND_UP((VMALLOC_END - VMALLOC_START), SZ_1M),
+		PAGE_OFFSET, (unsigned long)high_memory,
+		DIV_ROUND_UP(((unsigned long)high_memory - PAGE_OFFSET), SZ_1M),
+		MODULES_VADDR, MODULES_END,
+		DIV_ROUND_UP((MODULES_END - MODULES_VADDR), SZ_1M),
+
+		__init_begin, __init_end,
+		DIV_ROUND_UP((__init_end - __init_begin), SZ_1K),
+		_stext, _etext,
+		DIV_ROUND_UP((_etext - _stext), SZ_1K),
+		_sdata, _edata,
+		DIV_ROUND_UP((_edata - _sdata), SZ_1K));
+
+	BUILD_BUG_ON(TASK_SIZE				> MODULES_VADDR);
+	BUG_ON(TASK_SIZE				> MODULES_VADDR);
+
+	if (PAGE_SIZE >= 16384 && num_physpages <= 128) {
+		/*
+		 * On a machine this small we won't get
+		 * anywhere without overcommit, so turn
+		 * it on by default.
+		 */
+		sysctl_overcommit_memory = OVERCOMMIT_ALWAYS;
+	}
+}
+
+void free_initmem(void)
+{
+	totalram_pages += free_area(__phys_to_pfn(__pa(__init_begin)),
+				    __phys_to_pfn(__pa(__init_end)),
+				    "init");
+}
+
+#ifdef CONFIG_BLK_DEV_INITRD
+
+static int keep_initrd;
+
+void free_initrd_mem(unsigned long start, unsigned long end)
+{
+	if (!keep_initrd)
+		totalram_pages += free_area(__phys_to_pfn(__pa(start)),
+					    __phys_to_pfn(__pa(end)),
+					    "initrd");
+}
+
+static int __init keepinitrd_setup(char *__unused)
+{
+	keep_initrd = 1;
+	return 1;
+}
+
+__setup("keepinitrd", keepinitrd_setup);
+#endif
diff --git a/arch/unicore32/mm/ioremap.c b/arch/unicore32/mm/ioremap.c
new file mode 100644
index 00000000000..b7a605597b0
--- /dev/null
+++ b/arch/unicore32/mm/ioremap.c
@@ -0,0 +1,261 @@
+/*
+ * linux/arch/unicore32/mm/ioremap.c
+ *
+ * Code specific to PKUnity SoC and UniCore ISA
+ *
+ * Copyright (C) 2001-2010 GUAN Xue-tao
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ *
+ * Re-map IO memory to kernel address space so that we can access it.
+ *
+ * This allows a driver to remap an arbitrary region of bus memory into
+ * virtual space.  One should *only* use readl, writel, memcpy_toio and
+ * so on with such remapped areas.
+ *
+ * Because UniCore only has a 32-bit address space we can't address the
+ * whole of the (physical) PCI space at once.  PCI huge-mode addressing
+ * allows us to circumvent this restriction by splitting PCI space into
+ * two 2GB chunks and mapping only one at a time into processor memory.
+ * We use MMU protection domains to trap any attempt to access the bank
+ * that is not currently mapped.  (This isn't fully implemented yet.)
+ */
+#include <linux/module.h>
+#include <linux/errno.h>
+#include <linux/mm.h>
+#include <linux/vmalloc.h>
+#include <linux/io.h>
+
+#include <asm/cputype.h>
+#include <asm/cacheflush.h>
+#include <asm/mmu_context.h>
+#include <asm/pgalloc.h>
+#include <asm/tlbflush.h>
+#include <asm/sizes.h>
+
+#include <mach/map.h>
+#include "mm.h"
+
+/*
+ * Used by ioremap() and iounmap() code to mark (super)section-mapped
+ * I/O regions in vm_struct->flags field.
+ */
+#define VM_UNICORE_SECTION_MAPPING	0x80000000
+
+int ioremap_page(unsigned long virt, unsigned long phys,
+		 const struct mem_type *mtype)
+{
+	return ioremap_page_range(virt, virt + PAGE_SIZE, phys,
+				  __pgprot(mtype->prot_pte));
+}
+EXPORT_SYMBOL(ioremap_page);
+
+/*
+ * Section support is unsafe on SMP - If you iounmap and ioremap a region,
+ * the other CPUs will not see this change until their next context switch.
+ * Meanwhile, (eg) if an interrupt comes in on one of those other CPUs
+ * which requires the new ioremap'd region to be referenced, the CPU will
+ * reference the _old_ region.
+ *
+ * Note that get_vm_area_caller() allocates a guard 4K page, so we need to
+ * mask the size back to 4MB aligned or we will overflow in the loop below.
+ */
+static void unmap_area_sections(unsigned long virt, unsigned long size)
+{
+	unsigned long addr = virt, end = virt + (size & ~(SZ_4M - 1));
+	pgd_t *pgd;
+
+	flush_cache_vunmap(addr, end);
+	pgd = pgd_offset_k(addr);
+	do {
+		pmd_t pmd, *pmdp = pmd_offset((pud_t *)pgd, addr);
+
+		pmd = *pmdp;
+		if (!pmd_none(pmd)) {
+			/*
+			 * Clear the PMD from the page table, and
+			 * increment the kvm sequence so others
+			 * notice this change.
+			 *
+			 * Note: this is still racy on SMP machines.
+			 */
+			pmd_clear(pmdp);
+
+			/*
+			 * Free the page table, if there was one.
+			 */
+			if ((pmd_val(pmd) & PMD_TYPE_MASK) == PMD_TYPE_TABLE)
+				pte_free_kernel(&init_mm, pmd_page_vaddr(pmd));
+		}
+
+		addr += PGDIR_SIZE;
+		pgd++;
+	} while (addr < end);
+
+	flush_tlb_kernel_range(virt, end);
+}
+
+static int
+remap_area_sections(unsigned long virt, unsigned long pfn,
+		    size_t size, const struct mem_type *type)
+{
+	unsigned long addr = virt, end = virt + size;
+	pgd_t *pgd;
+
+	/*
+	 * Remove and free any PTE-based mapping, and
+	 * sync the current kernel mapping.
+	 */
+	unmap_area_sections(virt, size);
+
+	pgd = pgd_offset_k(addr);
+	do {
+		pmd_t *pmd = pmd_offset((pud_t *)pgd, addr);
+
+		set_pmd(pmd, __pmd(__pfn_to_phys(pfn) | type->prot_sect));
+		pfn += SZ_4M >> PAGE_SHIFT;
+		flush_pmd_entry(pmd);
+
+		addr += PGDIR_SIZE;
+		pgd++;
+	} while (addr < end);
+
+	return 0;
+}
+
+void __iomem *__uc32_ioremap_pfn_caller(unsigned long pfn,
+	unsigned long offset, size_t size, unsigned int mtype, void *caller)
+{
+	const struct mem_type *type;
+	int err;
+	unsigned long addr;
+	struct vm_struct *area;
+
+	/*
+	 * High mappings must be section aligned
+	 */
+	if (pfn >= 0x100000 && (__pfn_to_phys(pfn) & ~SECTION_MASK))
+		return NULL;
+
+	/*
+	 * Don't allow RAM to be mapped
+	 */
+	if (pfn_valid(pfn)) {
+		printk(KERN_WARNING "BUG: Your driver calls ioremap() on\n"
+			"system memory.  This leads to architecturally\n"
+			"unpredictable behaviour, and ioremap() will fail in\n"
+			"the next kernel release. Please fix your driver.\n");
+		WARN_ON(1);
+	}
+
+	type = get_mem_type(mtype);
+	if (!type)
+		return NULL;
+
+	/*
+	 * Page align the mapping size, taking account of any offset.
+	 */
+	size = PAGE_ALIGN(offset + size);
+
+	area = get_vm_area_caller(size, VM_IOREMAP, caller);
+	if (!area)
+		return NULL;
+	addr = (unsigned long)area->addr;
+
+	if (!((__pfn_to_phys(pfn) | size | addr) & ~PMD_MASK)) {
+		area->flags |= VM_UNICORE_SECTION_MAPPING;
+		err = remap_area_sections(addr, pfn, size, type);
+	} else
+		err = ioremap_page_range(addr, addr + size, __pfn_to_phys(pfn),
+					 __pgprot(type->prot_pte));
+
+	if (err) {
+		vunmap((void *)addr);
+		return NULL;
+	}
+
+	flush_cache_vmap(addr, addr + size);
+	return (void __iomem *) (offset + addr);
+}
+
+void __iomem *__uc32_ioremap_caller(unsigned long phys_addr, size_t size,
+	unsigned int mtype, void *caller)
+{
+	unsigned long last_addr;
+	unsigned long offset = phys_addr & ~PAGE_MASK;
+	unsigned long pfn = __phys_to_pfn(phys_addr);
+
+	/*
+	 * Don't allow wraparound or zero size
+	 */
+	last_addr = phys_addr + size - 1;
+	if (!size || last_addr < phys_addr)
+		return NULL;
+
+	return __uc32_ioremap_pfn_caller(pfn, offset, size, mtype, caller);
+}
+
+/*
+ * Remap an arbitrary physical address space into the kernel virtual
+ * address space. Needed when the kernel wants to access high addresses
+ * directly.
+ *
+ * NOTE! We need to allow non-page-aligned mappings too: we will obviously
+ * have to convert them into an offset in a page-aligned mapping, but the
+ * caller shouldn't need to know that small detail.
+ */
+void __iomem *
+__uc32_ioremap_pfn(unsigned long pfn, unsigned long offset, size_t size,
+		  unsigned int mtype)
+{
+	return __uc32_ioremap_pfn_caller(pfn, offset, size, mtype,
+			__builtin_return_address(0));
+}
+EXPORT_SYMBOL(__uc32_ioremap_pfn);
+
+void __iomem *
+__uc32_ioremap(unsigned long phys_addr, size_t size)
+{
+	return __uc32_ioremap_caller(phys_addr, size, MT_DEVICE,
+			__builtin_return_address(0));
+}
+EXPORT_SYMBOL(__uc32_ioremap);
+
+void __iomem *
+__uc32_ioremap_cached(unsigned long phys_addr, size_t size)
+{
+	return __uc32_ioremap_caller(phys_addr, size, MT_DEVICE_CACHED,
+			__builtin_return_address(0));
+}
+EXPORT_SYMBOL(__uc32_ioremap_cached);
+
+void __uc32_iounmap(volatile void __iomem *io_addr)
+{
+	void *addr = (void *)(PAGE_MASK & (unsigned long)io_addr);
+	struct vm_struct **p, *tmp;
+
+	/*
+	 * If this is a section based mapping we need to handle it
+	 * specially as the VM subsystem does not know how to handle
+	 * such a beast. We need the lock here b/c we need to clear
+	 * all the mappings before the area can be reclaimed
+	 * by someone else.
+	 */
+	write_lock(&vmlist_lock);
+	for (p = &vmlist ; (tmp = *p) ; p = &tmp->next) {
+		if ((tmp->flags & VM_IOREMAP) && (tmp->addr == addr)) {
+			if (tmp->flags & VM_UNICORE_SECTION_MAPPING) {
+				unmap_area_sections((unsigned long)tmp->addr,
+						    tmp->size);
+			}
+			break;
+		}
+	}
+	write_unlock(&vmlist_lock);
+
+	vunmap(addr);
+}
+EXPORT_SYMBOL(__uc32_iounmap);
diff --git a/arch/unicore32/mm/mm.h b/arch/unicore32/mm/mm.h
new file mode 100644
index 00000000000..3296bca0f1f
--- /dev/null
+++ b/arch/unicore32/mm/mm.h
@@ -0,0 +1,39 @@
+/*
+ * linux/arch/unicore32/mm/mm.h
+ *
+ * Code specific to PKUnity SoC and UniCore ISA
+ *
+ * Copyright (C) 2001-2010 GUAN Xue-tao
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+/* the upper-most page table pointer */
+extern pmd_t *top_pmd;
+extern int sysctl_overcommit_memory;
+
+#define TOP_PTE(x)	pte_offset_kernel(top_pmd, x)
+
+static inline pmd_t *pmd_off(pgd_t *pgd, unsigned long virt)
+{
+	return pmd_offset((pud_t *)pgd, virt);
+}
+
+static inline pmd_t *pmd_off_k(unsigned long virt)
+{
+	return pmd_off(pgd_offset_k(virt), virt);
+}
+
+struct mem_type {
+	unsigned int prot_pte;
+	unsigned int prot_l1;
+	unsigned int prot_sect;
+};
+
+const struct mem_type *get_mem_type(unsigned int type);
+
+extern void __flush_dcache_page(struct address_space *, struct page *);
+
+void __init bootmem_init(void);
+void uc32_mm_memblock_reserve(void);
diff --git a/arch/unicore32/mm/mmu.c b/arch/unicore32/mm/mmu.c
new file mode 100644
index 00000000000..7bf3d588631
--- /dev/null
+++ b/arch/unicore32/mm/mmu.c
@@ -0,0 +1,533 @@
+/*
+ * linux/arch/unicore32/mm/mmu.c
+ *
+ * Code specific to PKUnity SoC and UniCore ISA
+ *
+ * Copyright (C) 2001-2010 GUAN Xue-tao
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+#include <linux/module.h>
+#include <linux/kernel.h>
+#include <linux/errno.h>
+#include <linux/init.h>
+#include <linux/mman.h>
+#include <linux/nodemask.h>
+#include <linux/memblock.h>
+#include <linux/fs.h>
+#include <linux/bootmem.h>
+#include <linux/io.h>
+
+#include <asm/cputype.h>
+#include <asm/sections.h>
+#include <asm/setup.h>
+#include <asm/sizes.h>
+#include <asm/tlb.h>
+
+#include <mach/map.h>
+
+#include "mm.h"
+
+DEFINE_PER_CPU(struct mmu_gather, mmu_gathers);
+
+/*
+ * empty_zero_page is a special page that is used for
+ * zero-initialized data and COW.
+ */
+struct page *empty_zero_page;
+EXPORT_SYMBOL(empty_zero_page);
+
+/*
+ * The pmd table for the upper-most set of pages.
+ */
+pmd_t *top_pmd;
+
+pgprot_t pgprot_user;
+EXPORT_SYMBOL(pgprot_user);
+
+pgprot_t pgprot_kernel;
+EXPORT_SYMBOL(pgprot_kernel);
+
+static int __init noalign_setup(char *__unused)
+{
+	cr_alignment &= ~CR_A;
+	cr_no_alignment &= ~CR_A;
+	set_cr(cr_alignment);
+	return 1;
+}
+__setup("noalign", noalign_setup);
+
+void adjust_cr(unsigned long mask, unsigned long set)
+{
+	unsigned long flags;
+
+	mask &= ~CR_A;
+
+	set &= mask;
+
+	local_irq_save(flags);
+
+	cr_no_alignment = (cr_no_alignment & ~mask) | set;
+	cr_alignment = (cr_alignment & ~mask) | set;
+
+	set_cr((get_cr() & ~mask) | set);
+
+	local_irq_restore(flags);
+}
+
+struct map_desc {
+	unsigned long virtual;
+	unsigned long pfn;
+	unsigned long length;
+	unsigned int type;
+};
+
+#define PROT_PTE_DEVICE		(PTE_PRESENT | PTE_YOUNG |	\
+				PTE_DIRTY | PTE_READ | PTE_WRITE)
+#define PROT_SECT_DEVICE	(PMD_TYPE_SECT | PMD_PRESENT |	\
+				PMD_SECT_READ | PMD_SECT_WRITE)
+
+static struct mem_type mem_types[] = {
+	[MT_DEVICE] = {		  /* Strongly ordered */
+		.prot_pte	= PROT_PTE_DEVICE,
+		.prot_l1	= PMD_TYPE_TABLE | PMD_PRESENT,
+		.prot_sect	= PROT_SECT_DEVICE,
+	},
+	/*
+	 * MT_KUSER: pte for vecpage -- cacheable,
+	 *       and sect for unigfx mmap -- noncacheable
+	 */
+	[MT_KUSER] = {
+		.prot_pte  = PTE_PRESENT | PTE_YOUNG | PTE_DIRTY |
+				PTE_CACHEABLE | PTE_READ | PTE_EXEC,
+		.prot_l1   = PMD_TYPE_TABLE | PMD_PRESENT,
+		.prot_sect = PROT_SECT_DEVICE,
+	},
+	[MT_HIGH_VECTORS] = {
+		.prot_pte  = PTE_PRESENT | PTE_YOUNG | PTE_DIRTY |
+				PTE_CACHEABLE | PTE_READ | PTE_WRITE |
+				PTE_EXEC,
+		.prot_l1   = PMD_TYPE_TABLE | PMD_PRESENT,
+	},
+	[MT_MEMORY] = {
+		.prot_pte  = PTE_PRESENT | PTE_YOUNG | PTE_DIRTY |
+				PTE_WRITE | PTE_EXEC,
+		.prot_l1   = PMD_TYPE_TABLE | PMD_PRESENT,
+		.prot_sect = PMD_TYPE_SECT | PMD_PRESENT | PMD_SECT_CACHEABLE |
+				PMD_SECT_READ | PMD_SECT_WRITE | PMD_SECT_EXEC,
+	},
+	[MT_ROM] = {
+		.prot_sect = PMD_TYPE_SECT | PMD_PRESENT | PMD_SECT_CACHEABLE |
+				PMD_SECT_READ,
+	},
+};
+
+const struct mem_type *get_mem_type(unsigned int type)
+{
+	return type < ARRAY_SIZE(mem_types) ? &mem_types[type] : NULL;
+}
+EXPORT_SYMBOL(get_mem_type);
+
+/*
+ * Adjust the PMD section entries according to the CPU in use.
+ */
+static void __init build_mem_type_table(void)
+{
+	pgprot_user   = __pgprot(PTE_PRESENT | PTE_YOUNG | PTE_CACHEABLE);
+	pgprot_kernel = __pgprot(PTE_PRESENT | PTE_YOUNG |
+				 PTE_DIRTY | PTE_READ | PTE_WRITE |
+				 PTE_EXEC | PTE_CACHEABLE);
+}
+
+#define vectors_base()	(vectors_high() ? 0xffff0000 : 0)
+
+static void __init *early_alloc(unsigned long sz)
+{
+	void *ptr = __va(memblock_alloc(sz, sz));
+	memset(ptr, 0, sz);
+	return ptr;
+}
+
+static pte_t * __init early_pte_alloc(pmd_t *pmd, unsigned long addr,
+		unsigned long prot)
+{
+	if (pmd_none(*pmd)) {
+		pte_t *pte = early_alloc(PTRS_PER_PTE * sizeof(pte_t));
+		__pmd_populate(pmd, __pa(pte) | prot);
+	}
+	BUG_ON(pmd_bad(*pmd));
+	return pte_offset_kernel(pmd, addr);
+}
+
+static void __init alloc_init_pte(pmd_t *pmd, unsigned long addr,
+				  unsigned long end, unsigned long pfn,
+				  const struct mem_type *type)
+{
+	pte_t *pte = early_pte_alloc(pmd, addr, type->prot_l1);
+	do {
+		set_pte(pte, pfn_pte(pfn, __pgprot(type->prot_pte)));
+		pfn++;
+	} while (pte++, addr += PAGE_SIZE, addr != end);
+}
+
+static void __init alloc_init_section(pgd_t *pgd, unsigned long addr,
+				      unsigned long end, unsigned long phys,
+				      const struct mem_type *type)
+{
+	pmd_t *pmd = pmd_offset((pud_t *)pgd, addr);
+
+	/*
+	 * Try a section mapping - end, addr and phys must all be aligned
+	 * to a section boundary.
+	 */
+	if (((addr | end | phys) & ~SECTION_MASK) == 0) {
+		pmd_t *p = pmd;
+
+		do {
+			set_pmd(pmd, __pmd(phys | type->prot_sect));
+			phys += SECTION_SIZE;
+		} while (pmd++, addr += SECTION_SIZE, addr != end);
+
+		flush_pmd_entry(p);
+	} else {
+		/*
+		 * No need to loop; pte's aren't interested in the
+		 * individual L1 entries.
+		 */
+		alloc_init_pte(pmd, addr, end, __phys_to_pfn(phys), type);
+	}
+}
+
+/*
+ * Create the page directory entries and any necessary
+ * page tables for the mapping specified by `md'.  We
+ * are able to cope here with varying sizes and address
+ * offsets, and we take full advantage of sections.
+ */
+static void __init create_mapping(struct map_desc *md)
+{
+	unsigned long phys, addr, length, end;
+	const struct mem_type *type;
+	pgd_t *pgd;
+
+	if (md->virtual != vectors_base() && md->virtual < TASK_SIZE) {
+		printk(KERN_WARNING "BUG: not creating mapping for "
+		       "0x%08llx at 0x%08lx in user region\n",
+		       __pfn_to_phys((u64)md->pfn), md->virtual);
+		return;
+	}
+
+	if ((md->type == MT_DEVICE || md->type == MT_ROM) &&
+	    md->virtual >= PAGE_OFFSET && md->virtual < VMALLOC_END) {
+		printk(KERN_WARNING "BUG: mapping for 0x%08llx at 0x%08lx "
+		       "overlaps vmalloc space\n",
+		       __pfn_to_phys((u64)md->pfn), md->virtual);
+	}
+
+	type = &mem_types[md->type];
+
+	addr = md->virtual & PAGE_MASK;
+	phys = (unsigned long)__pfn_to_phys(md->pfn);
+	length = PAGE_ALIGN(md->length + (md->virtual & ~PAGE_MASK));
+
+	if (type->prot_l1 == 0 && ((addr | phys | length) & ~SECTION_MASK)) {
+		printk(KERN_WARNING "BUG: map for 0x%08lx at 0x%08lx can not "
+		       "be mapped using pages, ignoring.\n",
+		       __pfn_to_phys(md->pfn), addr);
+		return;
+	}
+
+	pgd = pgd_offset_k(addr);
+	end = addr + length;
+	do {
+		unsigned long next = pgd_addr_end(addr, end);
+
+		alloc_init_section(pgd, addr, next, phys, type);
+
+		phys += next - addr;
+		addr = next;
+	} while (pgd++, addr != end);
+}
+
+static void * __initdata vmalloc_min = (void *)(VMALLOC_END - SZ_128M);
+
+/*
+ * vmalloc=size forces the vmalloc area to be exactly 'size'
+ * bytes. This can be used to increase (or decrease) the vmalloc
+ * area - the default is 128m.
+ */
+static int __init early_vmalloc(char *arg)
+{
+	unsigned long vmalloc_reserve = memparse(arg, NULL);
+
+	if (vmalloc_reserve < SZ_16M) {
+		vmalloc_reserve = SZ_16M;
+		printk(KERN_WARNING
+			"vmalloc area too small, limiting to %luMB\n",
+			vmalloc_reserve >> 20);
+	}
+
+	if (vmalloc_reserve > VMALLOC_END - (PAGE_OFFSET + SZ_32M)) {
+		vmalloc_reserve = VMALLOC_END - (PAGE_OFFSET + SZ_32M);
+		printk(KERN_WARNING
+			"vmalloc area is too big, limiting to %luMB\n",
+			vmalloc_reserve >> 20);
+	}
+
+	vmalloc_min = (void *)(VMALLOC_END - vmalloc_reserve);
+	return 0;
+}
+early_param("vmalloc", early_vmalloc);
+
+static phys_addr_t lowmem_limit __initdata = SZ_1G;
+
+static void __init sanity_check_meminfo(void)
+{
+	int i, j;
+
+	lowmem_limit = __pa(vmalloc_min - 1) + 1;
+	memblock_set_current_limit(lowmem_limit);
+
+	for (i = 0, j = 0; i < meminfo.nr_banks; i++) {
+		struct membank *bank = &meminfo.bank[j];
+		*bank = meminfo.bank[i];
+		j++;
+	}
+	meminfo.nr_banks = j;
+}
+
+static inline void prepare_page_table(void)
+{
+	unsigned long addr;
+	phys_addr_t end;
+
+	/*
+	 * Clear out all the mappings below the kernel image.
+	 */
+	for (addr = 0; addr < MODULES_VADDR; addr += PGDIR_SIZE)
+		pmd_clear(pmd_off_k(addr));
+
+	for ( ; addr < PAGE_OFFSET; addr += PGDIR_SIZE)
+		pmd_clear(pmd_off_k(addr));
+
+	/*
+	 * Find the end of the first block of lowmem.
+	 */
+	end = memblock.memory.regions[0].base + memblock.memory.regions[0].size;
+	if (end >= lowmem_limit)
+		end = lowmem_limit;
+
+	/*
+	 * Clear out all the kernel space mappings, except for the first
+	 * memory bank, up to the end of the vmalloc region.
+	 */
+	for (addr = __phys_to_virt(end);
+	     addr < VMALLOC_END; addr += PGDIR_SIZE)
+		pmd_clear(pmd_off_k(addr));
+}
+
+/*
+ * Reserve the special regions of memory
+ */
+void __init uc32_mm_memblock_reserve(void)
+{
+	/*
+	 * Reserve the page tables.  These are already in use,
+	 * and can only be in node 0.
+	 */
+	memblock_reserve(__pa(swapper_pg_dir), PTRS_PER_PGD * sizeof(pgd_t));
+
+#ifdef CONFIG_PUV3_UNIGFX
+	/*
+	 * These should likewise go elsewhere.  They pre-reserve the
+	 * screen/video memory region at the 48M~64M of main system memory.
+	 */
+	memblock_reserve(PKUNITY_UNIGFX_MMAP_BASE, PKUNITY_UNIGFX_MMAP_SIZE);
+	memblock_reserve(PKUNITY_UVC_MMAP_BASE, PKUNITY_UVC_MMAP_SIZE);
+#endif
+}
+
+/*
+ * Set up device the mappings.  Since we clear out the page tables for all
+ * mappings above VMALLOC_END, we will remove any debug device mappings.
+ * This means you have to be careful how you debug this function, or any
+ * called function.  This means you can't use any function or debugging
+ * method which may touch any device, otherwise the kernel _will_ crash.
+ */
+static void __init devicemaps_init(void)
+{
+	struct map_desc map;
+	unsigned long addr;
+	void *vectors;
+
+	/*
+	 * Allocate the vector page early.
+	 */
+	vectors = early_alloc(PAGE_SIZE);
+
+	for (addr = VMALLOC_END; addr; addr += PGDIR_SIZE)
+		pmd_clear(pmd_off_k(addr));
+
+	/*
+	 * Create a mapping for UniGFX VRAM
+	 */
+#ifdef CONFIG_PUV3_UNIGFX
+	map.pfn = __phys_to_pfn(PKUNITY_UNIGFX_MMAP_BASE);
+	map.virtual = KUSER_UNIGFX_BASE;
+	map.length = PKUNITY_UNIGFX_MMAP_SIZE;
+	map.type = MT_KUSER;
+	create_mapping(&map);
+#endif
+
+	/*
+	 * Create a mapping for the machine vectors at the high-vectors
+	 * location (0xffff0000).  If we aren't using high-vectors, also
+	 * create a mapping at the low-vectors virtual address.
+	 */
+	map.pfn = __phys_to_pfn(virt_to_phys(vectors));
+	map.virtual = VECTORS_BASE;
+	map.length = PAGE_SIZE;
+	map.type = MT_HIGH_VECTORS;
+	create_mapping(&map);
+
+	/*
+	 * Create a mapping for the kuser page at the special
+	 * location (0xbfff0000) to the same vectors location.
+	 */
+	map.pfn = __phys_to_pfn(virt_to_phys(vectors));
+	map.virtual = KUSER_VECPAGE_BASE;
+	map.length = PAGE_SIZE;
+	map.type = MT_KUSER;
+	create_mapping(&map);
+
+	/*
+	 * Finally flush the caches and tlb to ensure that we're in a
+	 * consistent state wrt the writebuffer.  This also ensures that
+	 * any write-allocated cache lines in the vector page are written
+	 * back.  After this point, we can start to touch devices again.
+	 */
+	local_flush_tlb_all();
+	flush_cache_all();
+}
+
+static void __init map_lowmem(void)
+{
+	struct memblock_region *reg;
+
+	/* Map all the lowmem memory banks. */
+	for_each_memblock(memory, reg) {
+		phys_addr_t start = reg->base;
+		phys_addr_t end = start + reg->size;
+		struct map_desc map;
+
+		if (end > lowmem_limit)
+			end = lowmem_limit;
+		if (start >= end)
+			break;
+
+		map.pfn = __phys_to_pfn(start);
+		map.virtual = __phys_to_virt(start);
+		map.length = end - start;
+		map.type = MT_MEMORY;
+
+		create_mapping(&map);
+	}
+}
+
+/*
+ * paging_init() sets up the page tables, initialises the zone memory
+ * maps, and sets up the zero page, bad page and bad page tables.
+ */
+void __init paging_init(void)
+{
+	void *zero_page;
+
+	build_mem_type_table();
+	sanity_check_meminfo();
+	prepare_page_table();
+	map_lowmem();
+	devicemaps_init();
+
+	top_pmd = pmd_off_k(0xffff0000);
+
+	/* allocate the zero page. */
+	zero_page = early_alloc(PAGE_SIZE);
+
+	bootmem_init();
+
+	empty_zero_page = virt_to_page(zero_page);
+	__flush_dcache_page(NULL, empty_zero_page);
+}
+
+/*
+ * In order to soft-boot, we need to insert a 1:1 mapping in place of
+ * the user-mode pages.  This will then ensure that we have predictable
+ * results when turning the mmu off
+ */
+void setup_mm_for_reboot(char mode)
+{
+	unsigned long base_pmdval;
+	pgd_t *pgd;
+	int i;
+
+	/*
+	 * We need to access to user-mode page tables here. For kernel threads
+	 * we don't have any user-mode mappings so we use the context that we
+	 * "borrowed".
+	 */
+	pgd = current->active_mm->pgd;
+
+	base_pmdval = PMD_SECT_WRITE | PMD_SECT_READ | PMD_TYPE_SECT;
+
+	for (i = 0; i < FIRST_USER_PGD_NR + USER_PTRS_PER_PGD; i++, pgd++) {
+		unsigned long pmdval = (i << PGDIR_SHIFT) | base_pmdval;
+		pmd_t *pmd;
+
+		pmd = pmd_off(pgd, i << PGDIR_SHIFT);
+		set_pmd(pmd, __pmd(pmdval));
+		flush_pmd_entry(pmd);
+	}
+
+	local_flush_tlb_all();
+}
+
+/*
+ * Take care of architecture specific things when placing a new PTE into
+ * a page table, or changing an existing PTE.  Basically, there are two
+ * things that we need to take care of:
+ *
+ *  1. If PG_dcache_clean is not set for the page, we need to ensure
+ *     that any cache entries for the kernels virtual memory
+ *     range are written back to the page.
+ *  2. If we have multiple shared mappings of the same space in
+ *     an object, we need to deal with the cache aliasing issues.
+ *
+ * Note that the pte lock will be held.
+ */
+void update_mmu_cache(struct vm_area_struct *vma, unsigned long addr,
+	pte_t *ptep)
+{
+	unsigned long pfn = pte_pfn(*ptep);
+	struct address_space *mapping;
+	struct page *page;
+
+	if (!pfn_valid(pfn))
+		return;
+
+	/*
+	 * The zero page is never written to, so never has any dirty
+	 * cache lines, and therefore never needs to be flushed.
+	 */
+	page = pfn_to_page(pfn);
+	if (page == ZERO_PAGE(0))
+		return;
+
+	mapping = page_mapping(page);
+	if (!test_and_set_bit(PG_dcache_clean, &page->flags))
+		__flush_dcache_page(mapping, page);
+	if (mapping)
+		if (vma->vm_flags & VM_EXEC)
+			__flush_icache_all();
+}
diff --git a/arch/unicore32/mm/pgd.c b/arch/unicore32/mm/pgd.c
new file mode 100644
index 00000000000..08b8d4295e7
--- /dev/null
+++ b/arch/unicore32/mm/pgd.c
@@ -0,0 +1,102 @@
+/*
+ * linux/arch/unicore32/mm/pgd.c
+ *
+ * Code specific to PKUnity SoC and UniCore ISA
+ *
+ * Copyright (C) 2001-2010 GUAN Xue-tao
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+#include <linux/mm.h>
+#include <linux/gfp.h>
+#include <linux/highmem.h>
+
+#include <asm/pgalloc.h>
+#include <asm/page.h>
+#include <asm/tlbflush.h>
+
+#include "mm.h"
+
+#define FIRST_KERNEL_PGD_NR	(FIRST_USER_PGD_NR + USER_PTRS_PER_PGD)
+
+/*
+ * need to get a 4k page for level 1
+ */
+pgd_t *get_pgd_slow(struct mm_struct *mm)
+{
+	pgd_t *new_pgd, *init_pgd;
+	pmd_t *new_pmd, *init_pmd;
+	pte_t *new_pte, *init_pte;
+
+	new_pgd = (pgd_t *)__get_free_pages(GFP_KERNEL, 0);
+	if (!new_pgd)
+		goto no_pgd;
+
+	memset(new_pgd, 0, FIRST_KERNEL_PGD_NR * sizeof(pgd_t));
+
+	/*
+	 * Copy over the kernel and IO PGD entries
+	 */
+	init_pgd = pgd_offset_k(0);
+	memcpy(new_pgd + FIRST_KERNEL_PGD_NR, init_pgd + FIRST_KERNEL_PGD_NR,
+		       (PTRS_PER_PGD - FIRST_KERNEL_PGD_NR) * sizeof(pgd_t));
+
+	clean_dcache_area(new_pgd, PTRS_PER_PGD * sizeof(pgd_t));
+
+	if (!vectors_high()) {
+		/*
+		 * On UniCore, first page must always be allocated since it
+		 * contains the machine vectors.
+		 */
+		new_pmd = pmd_alloc(mm, (pud_t *)new_pgd, 0);
+		if (!new_pmd)
+			goto no_pmd;
+
+		new_pte = pte_alloc_map(mm, NULL, new_pmd, 0);
+		if (!new_pte)
+			goto no_pte;
+
+		init_pmd = pmd_offset((pud_t *)init_pgd, 0);
+		init_pte = pte_offset_map(init_pmd, 0);
+		set_pte(new_pte, *init_pte);
+		pte_unmap(init_pte);
+		pte_unmap(new_pte);
+	}
+
+	return new_pgd;
+
+no_pte:
+	pmd_free(mm, new_pmd);
+no_pmd:
+	free_pages((unsigned long)new_pgd, 0);
+no_pgd:
+	return NULL;
+}
+
+void free_pgd_slow(struct mm_struct *mm, pgd_t *pgd)
+{
+	pmd_t *pmd;
+	pgtable_t pte;
+
+	if (!pgd)
+		return;
+
+	/* pgd is always present and good */
+	pmd = pmd_off(pgd, 0);
+	if (pmd_none(*pmd))
+		goto free;
+	if (pmd_bad(*pmd)) {
+		pmd_ERROR(*pmd);
+		pmd_clear(pmd);
+		goto free;
+	}
+
+	pte = pmd_pgtable(*pmd);
+	pmd_clear(pmd);
+	pte_free(mm, pte);
+	pmd_free(mm, pmd);
+free:
+	free_pages((unsigned long) pgd, 0);
+}
diff --git a/arch/unicore32/mm/proc-macros.S b/arch/unicore32/mm/proc-macros.S
new file mode 100644
index 00000000000..51560d68c89
--- /dev/null
+++ b/arch/unicore32/mm/proc-macros.S
@@ -0,0 +1,145 @@
+/*
+ * linux/arch/unicore32/mm/proc-macros.S
+ *
+ * Code specific to PKUnity SoC and UniCore ISA
+ *
+ * Copyright (C) 2001-2010 GUAN Xue-tao
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ * We need constants.h for:
+ *  VMA_VM_MM
+ *  VMA_VM_FLAGS
+ *  VM_EXEC
+ */
+#include <generated/asm-offsets.h>
+#include <asm/thread_info.h>
+#include <asm/memory.h>
+
+/*
+ * the cache line sizes of the I and D cache are the same
+ */
+#define CACHE_LINESIZE	32
+
+/*
+ * This is the maximum size of an area which will be invalidated
+ * using the single invalidate entry instructions.  Anything larger
+ * than this, and we go for the whole cache.
+ *
+ * This value should be chosen such that we choose the cheapest
+ * alternative.
+ */
+#ifdef CONFIG_CPU_UCV2
+#define MAX_AREA_SIZE	0x800		/* 64 cache line */
+#endif
+
+/*
+ * vma_vm_mm - get mm pointer from vma pointer (vma->vm_mm)
+ */
+	.macro	vma_vm_mm, rd, rn
+	ldw	\rd, [\rn+], #VMA_VM_MM
+	.endm
+
+/*
+ * vma_vm_flags - get vma->vm_flags
+ */
+	.macro	vma_vm_flags, rd, rn
+	ldw	\rd, [\rn+], #VMA_VM_FLAGS
+	.endm
+
+	.macro	tsk_mm, rd, rn
+	ldw	\rd, [\rn+], #TI_TASK
+	ldw	\rd, [\rd+], #TSK_ACTIVE_MM
+	.endm
+
+/*
+ * act_mm - get current->active_mm
+ */
+	.macro	act_mm, rd
+	andn	\rd, sp, #8128
+	andn	\rd, \rd, #63
+	ldw	\rd, [\rd+], #TI_TASK
+	ldw	\rd, [\rd+], #TSK_ACTIVE_MM
+	.endm
+
+/*
+ * mmid - get context id from mm pointer (mm->context.id)
+ */
+	.macro	mmid, rd, rn
+	ldw	\rd, [\rn+], #MM_CONTEXT_ID
+	.endm
+
+/*
+ * mask_asid - mask the ASID from the context ID
+ */
+	.macro	asid, rd, rn
+	and	\rd, \rn, #255
+	.endm
+
+	.macro	crval, clear, mmuset, ucset
+	.word	\clear
+	.word	\mmuset
+	.endm
+
+#ifndef CONFIG_CPU_DCACHE_LINE_DISABLE
+/*
+ * va2pa va, pa, tbl, msk, off, err
+ *	This macro is used to translate virtual address to its physical address.
+ *
+ *	va: virtual address
+ *	pa: physical address, result is stored in this register
+ *	tbl, msk, off:	temp registers, will be destroyed
+ *	err: jump to error label if the physical address not exist
+ * NOTE: all regs must be different
+ */
+	.macro	va2pa, va, pa, tbl, msk, off, err=990f
+	movc	\pa, p0.c2, #0
+	mov	\off, \va >> #22		@ off <- index of 1st page table
+	adr	\tbl, 910f			@ tbl <- table of 1st page table
+900:						@ ---- handle 1, 2 page table
+	add	\pa, \pa, #PAGE_OFFSET		@ pa <- virt addr of page table
+	ldw	\pa, [\pa+], \off << #2		@ pa <- the content of pt
+	cand.a	\pa, #4				@ test exist bit
+	beq	\err				@ if not exist
+	and	\off, \pa, #3			@ off <- the last 2 bits
+	add	\tbl, \tbl, \off << #3		@ cmove table pointer
+	ldw	\msk, [\tbl+], #0		@ get the mask
+	ldw	pc, [\tbl+], #4
+930:						@ ---- handle 2nd page table
+	and	\pa, \pa, \msk			@ pa <- phys addr of 2nd pt
+	mov	\off, \va << #10
+	cntlo	\tbl, \msk			@ use tbl as temp reg
+	mov	\off, \off >> \tbl
+	mov	\off, \off >> #2		@ off <- index of 2nd pt
+	adr	\tbl, 920f			@ tbl <- table of 2nd pt
+	b	900b
+910:						@ 1st level page table
+	.word	0xfffff000, 930b		@ second level page table
+	.word	0xfffffc00, 930b		@ second level large page table
+	.word	0x00000000, \err		@ invalid
+	.word	0xffc00000, 980f		@ super page
+
+920:						@ 2nd level page table
+	.word	0xfffff000, 980f		@ page
+	.word	0xffffc000, 980f		@ middle page
+	.word	0xffff0000, 980f		@ large page
+	.word	0x00000000, \err		@ invalid
+980:
+	andn	\tbl, \va, \msk
+	and	\pa, \pa, \msk
+	or	\pa, \pa, \tbl
+990:
+	.endm
+#endif
+
+	.macro dcacheline_flush, addr, t1, t2
+	mov	\t1, \addr << #20
+	ldw	\t2, =_stext			@ _stext must ALIGN(4096)
+	add	\t2, \t2, \t1 >> #20
+	ldw	\t1, [\t2+], #0x0000
+	ldw	\t1, [\t2+], #0x1000
+	ldw	\t1, [\t2+], #0x2000
+	ldw	\t1, [\t2+], #0x3000
+	.endm
diff --git a/arch/unicore32/mm/proc-syms.c b/arch/unicore32/mm/proc-syms.c
new file mode 100644
index 00000000000..f30071e3665
--- /dev/null
+++ b/arch/unicore32/mm/proc-syms.c
@@ -0,0 +1,23 @@
+/*
+ * linux/arch/unicore32/mm/proc-syms.c
+ *
+ * Code specific to PKUnity SoC and UniCore ISA
+ *
+ * Copyright (C) 2001-2010 GUAN Xue-tao
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+#include <linux/module.h>
+#include <linux/mm.h>
+
+#include <asm/cacheflush.h>
+#include <asm/tlbflush.h>
+#include <asm/page.h>
+
+EXPORT_SYMBOL(cpu_dcache_clean_area);
+EXPORT_SYMBOL(cpu_set_pte);
+
+EXPORT_SYMBOL(__cpuc_dma_flush_range);
+EXPORT_SYMBOL(__cpuc_dma_clean_range);
diff --git a/arch/unicore32/mm/proc-ucv2.S b/arch/unicore32/mm/proc-ucv2.S
new file mode 100644
index 00000000000..9d296092e36
--- /dev/null
+++ b/arch/unicore32/mm/proc-ucv2.S
@@ -0,0 +1,134 @@
+/*
+ * linux/arch/unicore32/mm/proc-ucv2.S
+ *
+ * Code specific to PKUnity SoC and UniCore ISA
+ *
+ * Copyright (C) 2001-2010 GUAN Xue-tao
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+#include <linux/init.h>
+#include <linux/linkage.h>
+#include <asm/assembler.h>
+#include <asm/hwcap.h>
+#include <asm/pgtable-hwdef.h>
+#include <asm/pgtable.h>
+
+#include "proc-macros.S"
+
+ENTRY(cpu_proc_fin)
+	stm.w	(lr), [sp-]
+	mov	ip, #PSR_R_BIT | PSR_I_BIT | PRIV_MODE
+	mov.a	asr, ip
+	b.l	__cpuc_flush_kern_all
+	ldm.w	(pc), [sp]+
+
+/*
+ *	cpu_reset(loc)
+ *
+ *	Perform a soft reset of the system.  Put the CPU into the
+ *	same state as it would be if it had been reset, and branch
+ *	to what would be the reset vector.
+ *
+ *	- loc   - location to jump to for soft reset
+ */
+	.align	5
+ENTRY(cpu_reset)
+	mov	ip, #0
+	movc	p0.c5, ip, #28			@ Cache invalidate all
+	nop8
+
+	movc	p0.c6, ip, #6			@ TLB invalidate all
+	nop8
+
+	movc	ip, p0.c1, #0			@ ctrl register
+	or	ip, ip, #0x2000			@ vector base address
+	andn	ip, ip, #0x000f			@ ............idam
+	movc	p0.c1, ip, #0			@ disable caches and mmu
+	nop
+	mov	pc, r0				@ jump to loc
+	nop8
+
+/*
+ *	cpu_do_idle()
+ *
+ *	Idle the processor (eg, wait for interrupt).
+ *
+ *	IRQs are already disabled.
+ */
+ENTRY(cpu_do_idle)
+	mov	r0, #0				@ PCI address
+	.rept	8
+	ldw	r1, [r0]
+	.endr
+	mov	pc, lr
+
+ENTRY(cpu_dcache_clean_area)
+#ifndef CONFIG_CPU_DCACHE_LINE_DISABLE
+	csub.a	r1, #MAX_AREA_SIZE
+	bsg	101f
+	mov	r9, #PAGE_SZ
+	sub	r9, r9, #1			@ PAGE_MASK
+1:	va2pa	r0, r10, r11, r12, r13		@ r10 is PA
+	b	3f
+2:	cand.a	r0, r9
+	beq	1b
+3:	movc	p0.c5, r10, #11			@ clean D entry
+	nop8
+	add	r0, r0, #CACHE_LINESIZE
+	add	r10, r10, #CACHE_LINESIZE
+	sub.a	r1, r1, #CACHE_LINESIZE
+	bua	2b
+	mov	pc, lr
+#endif
+101:	mov	ip, #0
+	movc	p0.c5, ip, #10			@ Dcache clean all
+	nop8
+
+	mov	pc, lr
+
+/*
+ *	cpu_do_switch_mm(pgd_phys)
+ *
+ *	Set the translation table base pointer to be pgd_phys
+ *
+ *	- pgd_phys - physical address of new pgd
+ *
+ *	It is assumed that:
+ *	- we are not using split page tables
+ */
+	.align	5
+ENTRY(cpu_do_switch_mm)
+	movc	p0.c2, r0, #0			@ update page table ptr
+	nop8
+
+	movc	p0.c6, ip, #6			@ TLB invalidate all
+	nop8
+
+	mov	pc, lr
+
+/*
+ *	cpu_set_pte(ptep, pte)
+ *
+ *	Set a level 2 translation table entry.
+ *
+ *	- ptep  - pointer to level 2 translation table entry
+ *	- pte   - PTE value to store
+ */
+	.align	5
+ENTRY(cpu_set_pte)
+	stw	r1, [r0]
+#ifndef CONFIG_CPU_DCACHE_LINE_DISABLE
+	sub	r2, r0, #PAGE_OFFSET
+	movc	p0.c5, r2, #11				@ Dcache clean line
+	nop8
+#else
+	mov	ip, #0
+	movc	p0.c5, ip, #10				@ Dcache clean all
+	nop8
+	@dcacheline_flush	r0, r2, ip
+#endif
+	mov	pc, lr
+
diff --git a/arch/unicore32/mm/tlb-ucv2.S b/arch/unicore32/mm/tlb-ucv2.S
new file mode 100644
index 00000000000..061d455f9a1
--- /dev/null
+++ b/arch/unicore32/mm/tlb-ucv2.S
@@ -0,0 +1,89 @@
+/*
+ * linux/arch/unicore32/mm/tlb-ucv2.S
+ *
+ * Code specific to PKUnity SoC and UniCore ISA
+ *
+ * Copyright (C) 2001-2010 GUAN Xue-tao
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+#include <linux/init.h>
+#include <linux/linkage.h>
+#include <asm/assembler.h>
+#include <asm/page.h>
+#include <asm/tlbflush.h>
+#include "proc-macros.S"
+
+/*
+ *	__cpu_flush_user_tlb_range(start, end, vma)
+ *
+ *	Invalidate a range of TLB entries in the specified address space.
+ *
+ *	- start - start address (may not be aligned)
+ *	- end   - end address (exclusive, may not be aligned)
+ *	- vma   - vma_struct describing address range
+ */
+ENTRY(__cpu_flush_user_tlb_range)
+#ifndef	CONFIG_CPU_TLB_SINGLE_ENTRY_DISABLE
+	mov	r0, r0 >> #PAGE_SHIFT		@ align address
+	mov	r0, r0 << #PAGE_SHIFT
+	vma_vm_flags r2, r2			@ get vma->vm_flags
+1:
+	movc	p0.c6, r0, #3
+	nop8
+
+	cand.a	r2, #VM_EXEC			@ Executable area ?
+	beq	2f
+
+	movc	p0.c6, r0, #5
+	nop8
+2:
+	add	r0, r0, #PAGE_SZ
+	csub.a	r0, r1
+	beb	1b
+#else
+	movc	p0.c6, r0, #2
+	nop8
+
+	cand.a	r2, #VM_EXEC			@ Executable area ?
+	beq	2f
+
+	movc	p0.c6, r0, #4
+	nop8
+2:
+#endif
+	mov	pc, lr
+
+/*
+ *	__cpu_flush_kern_tlb_range(start,end)
+ *
+ *	Invalidate a range of kernel TLB entries
+ *
+ *	- start - start address (may not be aligned)
+ *	- end   - end address (exclusive, may not be aligned)
+ */
+ENTRY(__cpu_flush_kern_tlb_range)
+#ifndef	CONFIG_CPU_TLB_SINGLE_ENTRY_DISABLE
+	mov	r0, r0 >> #PAGE_SHIFT		@ align address
+	mov	r0, r0 << #PAGE_SHIFT
+1:
+	movc	p0.c6, r0, #3
+	nop8
+
+	movc	p0.c6, r0, #5
+	nop8
+
+	add	r0, r0, #PAGE_SZ
+	csub.a	r0, r1
+	beb	1b
+#else
+	movc	p0.c6, r0, #2
+	nop8
+
+	movc	p0.c6, r0, #4
+	nop8
+#endif
+	mov	pc, lr
+